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
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 Myanmar soldier guards an area at the Sittwe airport in Rakhine state, September 2018. Ye Aung THU / POOL / AFP
Report 325 / Asia

Avoiding a Return to War in Myanmar’s Rakhine State

An unofficial ceasefire has kept Rakhine State quiet compared to much of Myanmar following the 2021 coup. But friction is building between the military and ethnic Rakhine fighters. The parties should strike a formal deal to avert a return to war.

What’s new? After an informal ceasefire in late 2020, the Arakan Army used the lull in fighting to consolidate control of much of central and northern Rakhine State. Distracted by fallout from the 2021 coup, Myanmar’s military did little to oppose it at first, but rising tensions may lead to renewed combat.

Why does it matter? Many Rakhine State residents, including some Rohingya, have welcomed the shift to Arakan Army control, but the situation remains fraught. The ceasefire is fragile and the Arakan Army has grown significantly more powerful over the past eighteen months. A return to conflict would have devastating consequences for everyone in Rakhine.

What should be done? The Myanmar military and Arakan Army should avoid provoking a new war and formalise their ceasefire instead. The Arakan Army should eschew restrictions on humanitarian aid organisations, which should better coordinate their dealings with the group. Naypyitaw and Dhaka should open dialogue with the Arakan Army on Rohingya repatriation.

Executive Summary

Rakhine State has avoided the violence that has engulfed the rest of Myanmar since the February 2021 coup. The quiet owes in part to an informal ceasefire, which ended two years of fighting between the military and the Arakan Army, a pro-Rakhine ethnic armed group, and which came into force a few months before the military seized power in Naypyitaw. The Arakan Army has spurned the growing de facto alliance between the National Unity Government (NUG)-led opposition and other ethnic armed groups, focusing on getting control of much of Rakhine State. Until recently, the military has been too distracted to try loosening the Arakan Army’s grip, but tensions have started rising. The parties could soon find themselves back in conflict. While each side has reason to be leery of a formal ceasefire, both would also have reason to welcome the breathing space it would create. Most importantly, Rakhine’s people would benefit. In parallel, the Arakan Army should rein in demands on humanitarian actors, which should coordinate their interaction with it, and Dhaka and Naypyitaw should engage the group on Rohingya repatriation.

The two-year war that engulfed Rakhine State between late 2018 and 2020 significantly eroded Naypyitaw’s control of the region. Police and many other civil servants were often reluctant to leave major towns during the conflict due to the risk of attack or abduction and they remain wary of venturing into the countryside. Many local administrators resigned during this period due to threats either from the armed group or from the Myanmar military, which suspected them of collaborating with the enemy. The Arakan Army has since either replaced them with its own administrators in the areas it controls or co-opted the new appointees that the military regime has sent since the coup. As a result, the Arakan Army now directly or indirectly controls most rural areas in the centre and north of the state, while exerting significant influence in urban areas; it has also begun expanding farther south, as well as north toward the border with Bangladesh.

Over the past year, the Arakan Army has further consolidated its control by rolling out a suite of public services through its administrative branch, the Arakan People’s Authority. These include a judicial system and police force, which operate parallel to the state’s, and some health-care services (including the provision of COVID-19 vaccines). As a result, many residents are turning to Arakan Army mechanisms, rather than those run by Naypyitaw, for basic services and to resolve disputes. The service provision strategy has deepened public support for the group and its governance, but it is not without risk: it could be a major drain on the armed group’s resources, harm its popularity if the services do not live up to expectations and attract pushback from Naypyitaw.

The impact of these developments in Rakhine State has not been limited to the ethnic Rakhine community. The rise of the Arakan Army has brought positive changes for some hitherto ostracised Rohingya. While the overall situation for the Rohingya remains dire, some communities have improved access to public services and some are enjoying greater freedom of movement because of the Arakan Army’s non-enforcement of restrictions imposed by Naypyitaw. These testimonials should be considered in context, however; although Rohingya sources Crisis Group spoke to largely praised the Arakan Army and its administration, the community as a whole remains vulnerable and its members are generally not in a position to criticise the group for fear of reprisal.

Myanmar’s military regime … made only token efforts to counter the Arakan Army’s expanding control.

Myanmar’s military regime, which calls itself the State Administration Council, is focused on subduing resistance to the coup elsewhere in Myanmar and until recently made only token efforts to counter the Arakan Army’s expanding control. Part of the reason may be that locally based government and military officials, hunkered down in large towns, have little choice but to accept the facts on the ground. As a measure of the group’s growing influence, many state-run schools, which are still nominally under Naypyitaw’s control, have started playing an Arakan Army-written Rakhine anthem instead of the national anthem. There are even examples of active collaboration, such as Naypyitaw-controlled police working with the Arakan Army to resolve crimes and administrators from the two sides holding regular informal consultations.

But cooperation is certainly not the state’s default posture. The junta has in some instances sought to scare both Rakhine and Rohingya communities away from working with Arakan Army mechanisms and institutions. Recently, it has adopted more aggressive tactics, setting up roadblocks and searching vehicles, reinforcing troops, increasing patrols and detaining people it suspects of supporting the group.

The military is still too stretched to give much attention to Rakhine State, but there is a clear risk of a return to conflict. If the Arakan Army seeks to expand its influence consistent with its ambitious political aspirations – for example, into border areas or in southern Rakhine – it risks provoking the Myanmar military, which refers to itself as the Tatmadaw, into action. Similarly, a partnership with the NUG-led opposition – something many in Myanmar would welcome – could spark a return to war. While it would be difficult for the Tatmadaw to win this fight, the collateral effects of renewed conflict could be terrible for Rakhine State’s population, which is already reeling from neglect, a poor economy, communal violence and the earlier two-year war.

Although there is no clear path toward peace and stability for Rakhine State, one step that could offer both the parties and the region an extended respite from fighting might be to formalise the informal ceasefire that has largely kept the peace for the past eighteen months. Such an arrangement would focus primarily on maintaining the peace, in particular by demarcating territory and establishing formal communication channels to help de-escalate in the event that tensions begin to build once again.

Both the government and the Arakan Army have reason to be wary of such a step in that it would give the other party a chance to gather strength and prepare for renewed confrontation down the road. But each also has reason to embrace it. The military is preoccupied with the spiralling consequences of the coup that it launched over a year ago; a formal ceasefire would be a measure of insurance that it will not face another conflict that will stretch it further. As for the Arakan Army, such an arrangement could allow it to further consolidate its authority over the territories already under its control and gain recognition from outside actors, including both humanitarian organisations and neighbouring Bangladesh. While the extended respite could give the parties the chance to fortify themselves for further clashes, it would also create the possibility of a durable, peaceful solution emerging in the future – something outside actors could encourage.

If the parties agree to solidify the current ceasefire into a formal agreement, it will also be important to improve certain other ad hoc arrangements with implications for people in Rakhine. For example, humanitarian organisations are increasingly concerned that they may soon face parallel sets of requirements to operate in Rakhine State – some imposed by Naypyitaw and others by the Arakan Army – which could create both administrative burdens and operational difficulties. The group should work to ensure that those in need are not cut off from humanitarian assistance because of paralysing new rules and humanitarian organisations should come together to present a united front should such rules become overly burdensome. Dhaka and Naypyitaw should additionally engage with the Arakan Army on the possible return of Rohingya refugees living in Bangladesh.

While a formal ceasefire would offer neither party precisely what it wants, there would be enough in it for both sides that it could conceivably work. That in turn would allow Rakhine State residents to get what they need most: a continued break from violence and the corresponding opportunity to build toward a more peaceful future, one in which Rakhine and Rohingya can live in relative safety side by side.

Brussels, 1 June 2022

I. Introduction

Since independence in 1948, the Myanmar state has never been in full control of all the territory within its borders; at times, it has controlled little more than the major cities and key infrastructure, with the rest in dispute or in the hands of various non-state armed groups. Over the decades, some of these armed groups have created state-like enclaves in the country’s borderlands within which they provide services, issue laws, maintain law and order, collect taxes and do business. The United Wa State Army, the Kachin Independence Organisation and the Karen National Union are among those with well-developed governance systems, but even the smaller of Myanmar’s twenty-plus ethnic armed groups exert control over some territory.

Rakhine State has been an outlier. There, insurgency largely failed to take root and the state was firmly entrenched. Unlike most other minorities, the ethnic Rakhine attained high-ranking positions in both the military and the civil service, alongside the majority Burmans. But the Rakhine also harboured deep grievances toward the Burmans, dating back centuries to the conquest of the Arakan kingdom of Mrauk-U in 1784 by a Burman king, Bodawpaya. They pointed to Rakhine’s deep poverty as evidence of the Burman-dominated state’s neglect of the people of Rakhine, both Buddhist and Muslim.

Their anger, though, has instead often been directed toward the Rohingya, a Muslim minority in Rakhine. In 2012, communal conflict in the state left close to 200 people dead – mostly Muslims – and almost 150,000 displaced. The communities were segregated and tensions regularly threatened to boil over into further violence. Some Rakhine were implicated in the military’s deadly campaign against the Rohingya in August-September 2017 that sent more than 700,000 Muslims fleeing to Bangladesh.

State control had dissipated in much of the centre and north, leaving a vacuum that the Arakan Army set out to fill.

Against this backdrop, a group of young Rakhine exiles established the Arakan Army in 2009, with the support of the Kachin Independence Army, and quietly built up their forces in northern Myanmar. From 2014, they began inserting their troops into Rakhine State and spreading an ethno-nationalist ideology that shifted the blame for Rakhine’s woes back to Naypyitaw. A brutal war erupted in December 2018; when the two sides reached a surprise ceasefire in November 2020, state control had dissipated in much of the centre and north, leaving a vacuum that the Arakan Army set out to fill.

This report examines how the Arakan Army has used the February 2021 coup to cement control over much of Rakhine State and why Myanmar’s military regime has not taken decisive action to stop it. It looks at the impact this shift in control has had on the lives of both Rakhine and Rohingya residents, as well as the implications for humanitarian aid and the repatriation of Rohingya refugees in Bangladesh. It is based on research conducted between January and May 2022 and builds on Crisis Group’s years of fieldwork and analysis on conflict dynamics in Myanmar. Given the constraints on travel due to COVID-19 and the military takeover, research was conducted remotely, using pre-existing networks of contacts. Sources included Rohingya and Rakhine residents of Rakhine State, members of civil society organisations and NGOs, diplomats and aid workers, and analysts and individuals close to the key protagonists.[fn]For Crisis Group reporting on the rise of the Arakan Army, see Crisis Group Asia Briefings N°s 154, A New Dimension of Violence in Myanmar’s Rakhine State, 24 January 2019; 164, From Elections to Ceasefire in Myanmar’s Rakhine State, 23 December 2020; and Asia Report N°307, An Avoidable War: Politics and Armed Conflict in Myanmar’s Rakhine State, 9 June 2020.Hide Footnote

II. The Arakan Army: From Conflict to Control

A. A Productive Peace

After fighting erupted in Rakhine and southern Chin State in late 2018, the Arakan Army began systematically dismantling Naypyitaw’s administrative system through a campaign of violence and intimidation. Local administrators resigned en masse due to safety concerns, while township-level officials and other civil servants, including police, were often unwilling to travel outside of urban areas.[fn]Crisis Group Report, An Avoidable War, op. cit. See also, for example, “Local officials resign en masse in Myanmar’s conflict-torn Rakhine State”, The Irrawaddy, 22 June 2020.Hide Footnote  In December 2019, the group formed a wing called the Arakan People’s Authority to administer recently captured territory and began actively recruiting ethnic Rakhine with administrative experience from elsewhere in Myanmar (and even abroad).[fn]Crisis Group interview, Rakhine journalist close to the Arakan Army, September 2020. See also Crisis Group Report, An Avoidable War, op. cit.; “Arakan Army seizes ferry, unveils taxation agency, in Myanmar’s Rakhine State”, Radio Free Asia, 31 December 2019; and “Arakan Army collects taxes, polices streets in parts of Myanmar’s war-torn Rakhine State”, Radio Free Asia, 20 July 2020.Hide Footnote  But with war raging across much of the region, implementation proved challenging.

The November 2020 de facto ceasefire changed the equation.[fn]Crisis Group Report, From Elections to Ceasefire, op. cit.Hide Footnote  With the fighting on pause and the new military regime focused on subduing resistance to its rule in other parts of the country, the Arakan Army has been able to strengthen its grip upon much of the centre and north of the state.[fn]Myanmar Edges Toward State Collapse, 1 April 2021; 168, Taking Aim at the Tatmadaw: The New Armed Resistance to Myanmar’s Coup, 28 June 2021; 170, The Deadly Stalemate in Post-Coup Myanmar, 20 October 2021; and 171, Resisting the Resistance: Myanmar’s Pro-Military Pyusawhti Militias, 6 April 2022; Asia Reports N°s 314, Myanmar’s Military Struggles to Control the Virtual Battlefield, 18 May 2021 and 319, Myanmar’s Coup Shakes Up Its Ethnic Conflicts, 12 January 2022. See also Richard Horsey, “A Close-up View of Myanmar’s Leaderless Mass Protests”, Crisis Group Commentary, 26 February 2021 and Richard Horsey, “One Year On from the Myanmar Coup, Crisis Group Commentary, 25 January 2022.Hide Footnote It created new administrative boundaries and appointed political officers, senior administrators, judicial officers and police officers.[fn]The new administrative districts are known as Alpha, Victor and Nova. Within each of these, there are three township equivalents, known as Alpha-1, Alpha-2 and so on. Crisis Group interviews, various sources familiar with the situation in Rakhine State, February-March 2022.Hide Footnote  Strong support for the armed group among the majority Rakhine population made the task significantly easier, to the extent that it has received support from some of the civil servants on Naypyitaw’s payroll. The rapid and relatively successful rollout of services to rural communities has only further enhanced its legitimacy and popularity, particularly among ethnic Rakhine, even though this expansion has not been without teething problems.

The Arakan Army is also making a concerted effort to formally separate its military and political/administrative functions and to reduce the influence of military officers over administrative activities.[fn]Crisis Group interviews, Rakhine researcher and Rakhine journalist, February 2022.Hide Footnote  This attempt to transform itself from an army into a state-like entity, and the group’s growing role in the everyday lives of ordinary people in Rakhine State, is reflected in the local vernacular. Residents in Rakhine now generally use the name of the group’s political wing, the United League of Arakan (ULA) – which oversees the Arakan People’s Authority – rather than the Arakan Army, the armed wing, when referring to the group’s non-military activities. Within only a couple of years, the group’s governance apparatus has begun to resemble those of much more established ethnic armed groups that have long governed sizeable territory, such as the Karen National Union in south-eastern Myanmar and the Kachin Independence Organisation in northern Myanmar.

The [Arakan Army] now has de facto authority over somewhere from 50 to 75 per cent of the territory of Rakhine State.

Given the speed with which its administrative expansion has unfolded and the lack of territorial demarcation, it is difficult to pinpoint how much territory the Arakan Army actually controls. But it appears the insurgent group now has de facto authority over somewhere from 50 to 75 per cent of the territory of Rakhine State. Generally speaking, Naypyitaw still controls cities and most of southern Rakhine, particularly Gwa, Thandwe, Taungup and Munaung townships. In the north, rural areas of Mrauk-U, Kyauktaw, Rathedaung, Buthidaung and Ponnagyun townships are either under total Arakan Army control, with no remaining state structures, or under its sway in effect, with state structures co-opted to serve its agenda. The strategically important border areas in Maungdaw Township, as well as parts of central and southern Rakhine, such as Kyaukpyu, Taungup and Ann townships, are more contested, with competing power structures.[fn]Crisis Group interviews, various sources familiar with the situation in Rakhine State, January-March 2022.Hide Footnote

Notably, the group has not taken a strong position against the February 2021 military coup. Its leadership has actively discouraged anti-regime protests and civil disobedience, urging the public to focus on the goal of Rakhine autonomy while paying less attention to political developments elsewhere in the country.[fn]"AA chief does not want Myanmar’s strikes and protests in Rakhine State”, The Irrawaddy, 12 April 2021.Hide Footnote  Unlike some other ethnic armed groups, it has also not cooperated with the parallel National Unity Government (NUG) formed by lawmakers ousted by the coup, despite public invitations from the latter to join the struggle against the regime.[fn]See Crisis Group Report, Myanmar’s Coup Shakes Up Its Ethnic Conflicts, op. cit.; and “အမျိုးသားညီညွတ်ရေးအစိုးရ NUG ယာယီသမ္မတနှင့်ဇနီးတို့ အိမ်ထောင်သက် နှစ် ၄၀ ပြည့် မင်္ဂလာအထိမ်းအမှတ် AA က မင်္ဂလာဆုတောင်းသဝဏ်လွှာပို့ [Arakan Army sends congratulatory message to National Unity Government acting president and wife on 40th wedding anniversary]”, Khit Thit Media, 12 December 2021.Hide Footnote  The Arakan Army has also insisted that it is has no contact with the anti-coup militias (many of which are known as People’s Defence Forces) that have emerged in opposition to the regime.[fn]“Rebel yell: Arakan Army leader speaks to Asia Times”, Asia Times, 18 January 2022.Hide Footnote  As tensions with the military have increased in recent months, however, it has held informal talks with the NUG and publicly acknowledged training some anti-coup resistance forces.[fn]Crisis Group interview, source close to the Arakan Army, March 2022. When the Arakan Army marked its thirteenth anniversary on 10 April 2022, it published congratulatory statements from five resistance groups in which they thanked the Army for its training and support. Most of these groups are based in the mountain ranges on the edge of Rakhine State; the Arakan Army appears to be using them to create a buffer zone between its forces and the rest of the country, particularly Magway Region and Chin State.Hide Footnote

B. Acting Like a State

The Arakan Army’s state-building aspirations in Rakhine State long predate the November 2020 de facto ceasefire. Leader Twan Mrat Naing has described the group’s political goal as a “confederate” status for Rakhine State under which the territory would enjoy almost complete autonomy. Appealing to memories of Rakhine’s status as an independent kingdom, Twan Mrat Naing describes the Arakan Army’s revolutionary struggle as the “Way of Rakhita” and its goal as the “Arakan Dream”.[fn]The “Way of Rakhita” refers to the method – revolutionary struggle – of achieving the “Arakan Dream” of an independent or autonomous state. “Arakan” is the historical name for “Rakhine”, a term Myanmar’s military regime adopted in 1989. For a more detailed discussion, see Crisis Group Report, An Avoidable War, op. cit.; and Kyaw Lynn, “The Arakan Army, Myanmar military coup and politics of Arakan”, Transnational Institute, 10 June 2021.Hide Footnote  More recently, he has articulated phases of revolutionary struggle, including a state-building phase, and spoken of the need to anchor the movement in strong institutions rather than simply in ethno-nationalist ideology and its leaders’ popularity.[fn]“ရက္ခိုင်အမျိုးသားအဖွဲချုပ် ULA ဥက္ကဌ ရက္ခိုင့်တပ်တော်တပ်မှူးချုပ် ဗိုလ်ချုပ်ထွန်းမြတ်နိုင်နန့် အင်တာ [Interview with ULA chairman and AA commander-in-chief Twan Mrat Naing]”, Arrakha Media, 15 August 2021.Hide Footnote  He has also warned that the group will push for independence if “our rightful political status … is not accommodated within this union”.[fn]“Rebel yell”, op. cit.Hide Footnote

The Arakan Army’s first priority was to establish an administrative system for the areas it controls.

The Arakan People’s Authority is the embodiment of these aspirations. The body has a range of functions, but the Arakan Army’s first priority was to establish an administrative system for the areas it controls. The group moved quickly to fill the void left by the mass resignation of civil servants in 2019-2020, appointing its own administrators, who sit in offices under the ULA flag and often have little or no contact with the military regime. These local administrators head committees comprising community leaders who carry out a range of functions, including tax collection, dispute resolution and criminal investigations, and answer to higher-ranking administrators and political officers.[fn]Crisis Group interviews, various sources familiar with the situation in Rakhine State, January-March 2022.Hide Footnote

In many areas, though, the Arakan Army has instead co-opted incumbent administrators appointed since the coup by the General Administration Department under Myanmar’s Ministry of Home Affairs. Although they still nominally serve Naypyitaw, these administrators are in effect under the Arakan People’s Authority’s control, seeking its guidance in local decision-making and reporting back after meetings with their township superiors.[fn]Ibid.Hide Footnote  Because one of these local administrators’ responsibilities is to recommend how public funding should be allocated and oversee how it is spent, working through them allows the Arakan Army access to state funds to develop its newly gained territory, and thus reduce its own expenses. Similarly, these administrators have enabled local farmers in insurgent-held areas to obtain state agricultural loans.[fn]Crisis Group interview, researcher focused on Rakhine State, January 2022.Hide Footnote

The enlargement of the Arakan Army’s administrative footprint has enabled it to expand and formalise tax collection. Previously, the group collected taxes in an ad hoc manner, but since early 2021 it has nominally established fixed rates for both households and businesses. Households are expected to pay a flat tax, which is usually 3,000 kyat a month (about $1.50). Businesses, meanwhile, are supposed to pay between 2 and 5 per cent of their “total investment” on an annual basis; this term appears to refer to their capitalisation, but the precise means of calculating the amount are unclear and in practice it is usually just a negotiated sum. Sources indicate that, for now at least, households are generally willing to pay; some business owners are more reluctant, but fear of arrest means they usually hand over the money as well.[fn]A resident of Mrauk-U Township related that the Arakan Army detained a relative for refusing to pay a 5 per cent tax on their business, but later released him after he agreed to pay 3 per cent. Crisis Group interview, January 2022.Hide Footnote

While they are paying taxes, residents are also benefiting from new services, particularly improved law and order. With its administrative network in place, the Arakan Army has made judicial services and law enforcement a primary focus. The group was already fulfilling these roles to some extent prior to the November 2020 ceasefire by occasionally acting on complaints from the public to apprehend (and punish) alleged criminals.[fn]Crisis Group Report, An Avoidable War, op. cit.Hide Footnote  Over the past year, however, it has established both a judiciary and police force that are separate from its armed wing, falling under the Arakan People’s Authority. Disputes that cannot be resolved by village administration committees are referred to newly created ULA courts.[fn]Crisis Group interviews, Rakhine residents and journalist, January-March 2022.Hide Footnote

The judicial system, in particular, has seen high uptake among Rakhine State residents, who have long been frustrated at the corruption, partiality and inefficiency in the Naypyitaw-controlled judiciary. Cases that would have taken a year or more to resolve through the state system, and incurred significant expenses in bribes and fees, can now be resolved in as little as a month, at minimal expense.[fn]Crisis Group interviews, Rakhine State based researcher and political analyst, January 2022.Hide Footnote  Sources Crisis Group spoke to estimated that in areas where the Arakan Army system operates, over three fourths of criminal and non-criminal disputes are resolved through these mechanisms.[fn]Crisis Group interviews, residents of Mrauk-U, Sittwe and Ramree townships, January 2022.Hide Footnote  Many judicial officials have actually resigned to join the Arakan People’s Authority, providing it with some trained judges. As one civil society leader in Ramree Township noted:

The Arakan People’s Authority oversees around 90 per cent of criminal and non-criminal cases here – even some that happen in downtown Ramree. Only people related to the military or civil service use the junta courts.[fn]Crisis Group interview, resident of Ramree township, January 2022.Hide Footnote

C. Strategic Rollout

Because state structures still exist in most parts of Rakhine State, the Arakan Army has been able to pick which services it will offer and which it will leave to Naypyitaw. It has chosen in a strategic manner, allocating its resources for maximum effectiveness. Replacing or co-opting the state administrative apparatus is not only an important symbol of Arakan Army control, but also a means of depriving Naypyitaw of a physical presence on the ground and access to intelligence.[fn]For a detailed examination of the General Administration Department, see “Administering the State in Myanmar: An Overview of the General Administration Department, The Asia Foundation, March 2015.Hide Footnote  Tax collection flows from administrative control, providing the group with an important source of funding and boosting its legitimacy when it uses the money for effective service delivery in a kind of “virtuous circle”. Complaints about police and judicial corruption, and the general breakdown of law enforcement in many areas due to the 2019-2020 conflict, had created an obvious vacuum for the armed group to fill.

In contrast, the Arakan Army has let Naypyitaw continue to provide most health and education services in areas it controls. Both require significant financial and human resources, and providing them would likely be well beyond the armed group’s capacity. Nevertheless, its influence – and that of Rakhine nationalism more broadly – permeates the classroom. Over the past two years, the Rakhine national anthem, written by the Arakan Army, has replaced the Myanmar national anthem in government-run schools in many areas.[fn]Crisis Group interviews, various sources in Rakhine State, January-March 2022. See also “Some schools opt for Arakha national anthem over its Myanmar counterpart”, Development Media Group, 20 January 2022.Hide Footnote  The armed group does not appear to have forced the anthem change, but it has obviously inspired the switch with its Rakhine nationalist ideology. “We started singing it [the Rakhine anthem] at our school because we wanted students to know our identity”, said a teacher.[fn]Crisis Group interview, ethnic Rakhine state schoolteacher from Buthidaung Township, January 2022.Hide Footnote  In some places, though not as many, the ULA flag has also replaced the national flag.

The Arakan Army’s administrative and governance structures are far from unique in Myanmar. For decades, ethnic armed groups in the country’s borderlands have run their own schools, courts and administrative systems in territories under their control.[fn]See Crisis Group Asia Report N°52, Myanmar Backgrounder: Ethnic Minority Politics, 7 May 2003. See also Aung Naing Oo, “A cold war in Myanmar and the dangers of a protracted ceasefire”, Frontier Myanmar, 1 February 2019.Hide Footnote  These services have been important for building acceptance of these insurgent movements, creating what some have described as a “social contract” between their leaders and the population.[fn]For a detailed examination of this issue and how it has affected the Kachin and Karen insurgencies in recent decades, see David Brenner, Rebel Politics: A Political Sociology of Armed Struggle in Myanmar’s Borderlands (Ithaca, 2019). See also the work of Kim Jolliffe, including “Ethnic Conflict and Social Services in Myanmar’s Contested Regions”, The Asia Foundation, 2014.Hide Footnote  Although it has not yet seriously started to take on responsibility for health and education, the Arakan Army is thus following a well-worn path, and interviews with Rakhine State residents suggest that the group is indeed benefiting in terms of enhanced acceptance and stronger buy-in to its vision for Rakhine State. Significantly, this support is not limited to the majority Rakhine Buddhists, but can also be found to some degree among the state’s Rohingya Muslim population as well (see Section IV for more).[fn]Crisis Group interviews, Rakhine State residents, January-February 2022.Hide Footnote

D. The Risk of Overreach

The rapid expansion of the Arakan People’s Authority has inevitably created challenges for what was, until less than two years ago, an exclusively military organisation. Lack of human resources has been a chronic problem, with the group heavily relying on civil servants who have quit the state system or who are still working for Naypyitaw. Many policies and structures are still not formalised or uniformly enforced, as reflected in the differing tax rates from locality to locality. Unlike other ethnic armed groups, the Arakan Army has not yet developed its own laws or school curriculum. The judicial system also remains relatively informal, with disputes still often resolved through mediation. The high caseload and the pressure to deliver results means that the courts emphasise speed rather than procedure.[fn]Crisis Group interviews, various sources with knowledge of Rakhine State, January-March 2022.Hide Footnote

Since early 2022, criticism of the system’s shortcomings has emerged. Rakhine-based media organisations have reported claims of corruption and bias within the Arakan People’s Authority, particularly in the judicial system, and negative posts have started proliferating on social media.[fn]See, for example, “ရက္ခိုင်ပြည်သူ့အာဏာပိုင်အဖွဲ့ လက်အောက်ရှိ တရားရေးဌာနအချို့၏ အမှုဖြေရှင်းပေးမှုနှင့် တရားစီရင်မှုပိုင်းတို့၌ အမှုတချို့တွင် မကျေနပ်မှု ရှိနေ [Dissatisfaction with the handling of some cases by the Arakan People’s Authority judiciary]”, Western Media, 22 February 2021.Hide Footnote  In March 2022, the Arakan Army responded with the first of what it said would be regular monthly press conferences – an unusual move for an ethnic armed group in Myanmar – at which it acknowledged challenges in its governance system.[fn]Ahead of the press conference, the group issued five reasons for meeting journalists, one of which was to address criticism of its judiciary and to help the public “better understand the weaknesses” of its administrative system. “သတင်းစာရှင်းလင်းပွဲပြုလုပ်ရသည့် ရည်ရွယ်ချက်များ [Objectives of holding the press conference]”, ULA/AA Press Briefing Info, 5 March 2022.Hide Footnote  In particular, spokesman Khaing Thu Kha admitted that some officials had abused their powers and promised that the group would take action, before inviting those criticising the Arakan People’s Authority to join its ranks and help improve the civil service.[fn]“အုပ်ချုပ်ရေး၊ တရားစီရင်ရေးမဏ္ဍိုင်များတွင် ခေတ်ပညာတက်လူငယ်များပါဝင်ရန် ULA/AA ဖိတ်ခေါ် [ULA invites educated youth to join administration, judiciary]”, Western Media, 5 March 2022.Hide Footnote

Despite its rapid strides, the Arakan Army will face a constant battle to meet the expectations it has created among its constituents.

Despite its rapid strides, the Arakan Army will face a constant battle to meet the expectations it has created among its constituents. Particularly at a time of relative peace, it runs the risk of losing public support if its officials appear to underperform. So far, however, many Rakhine appear to be willing to overlook shortcomings; instead, they take pride in the Arakan Army’s achievements and express confidence in the group’s ability to improve the new system and, eventually, achieve its goal of autonomy or independence. A Buthidaung teacher complained, for instance, that when she was a student, she was “taught nonsense Burmese history at school”. Her students now sing the Rakhine national anthem, though for the moment she still teaches them using the same state curriculum she disparages.[fn]Crisis Group interview, Buthidaung teacher, January 2022.Hide Footnote  Nonetheless, the Arakan Army’s influence has already changed the way she does her job.

The Arakan Army’s administrative expansion could also place it under financial strain that could either hinder its activities or force it to take steps to increase its revenues. Both are fraught with risk. Present tax receipts are insufficient to support the group’s expansion, but levying higher taxes could create hardship and resentment among its supporters, particularly given high poverty levels in Rakhine State.[fn]Rakhine is Myanmar’s poorest state or region, with 78 per cent of the population below the poverty line, compared to a national average of 37.5 per cent. See “Myanmar: Ending Poverty and Boosting Shared Prosperity in a Time of Transition”, World Bank, November 2014.Hide Footnote  As for finding new sources of funding, the Arakan Army could well be tempted to venture further into illicit activities, as many armed groups in Myanmar (including the military and its allies) have done to sustain their operations. The Arakan Army has previously been accused of involvement in drug trafficking.[fn]Crisis Group Report, A New Dimension of Violence, op. cit.Hide Footnote  Deepening such involvement could harm its public image – indeed, the military has in the past used allegations of involvement in the drug trade to delegitimise the group. But Rakhine State offers only limited opportunities to pursue legitimate business opportunities or more accepted illicit activities, such as informal border crossings, or the taxing of timber or mining, as is common elsewhere in Myanmar.[fn]Crisis Group interview, source close to the State Administration Council, March 2022.Hide Footnote

Arakan Army leaders will likely be aware of the risks that accompany the current transition, given the experience of the Kachin Independence Organisation (KIO), which supported the Arakan Army’s creation and with which it maintains close ties.[fn]Crisis Group Reports, A New Dimension of Violence and An Avoidable War, both op. cit.Hide Footnote  The KIO underwent a similar transition from active guerrilla warfare to a focus on governance after signing a 1994 ceasefire with the military. In subsequent years, KIO leaders lost the support of the public and the group’s rank-and-file after engaging in business deals with military elites and regime cronies that made them and their families wealthy but brought few benefits to – and even often hurt – their Kachin constituency. They also failed to properly fund the group’s administrative functions, adding to community frustrations. Eventually, the KIO leadership that signed the ceasefire was overthrown. The new, younger generation of leaders rejected the Tatmadaw’s demands in 2009 to become a military-affiliated Border Guard Force, leading to the breakdown of the 1994 ceasefire two years later.[fn]See Crisis Group Asia Briefing N°140, A Tentative Peace in Myanmar’s Kachin Conflict, 12 June 2013, Section III. See also Brenner, op. cit., chapter 4.Hide Footnote

III. Naypyitaw’s Long Leash

The Myanmar military’s seizure of power in February 2021 has created ideal conditions for the Arakan Army’s state-building agenda.[fn]It is ironic that the military takeover would have this effect, given that the November 2020 de facto ceasefire likely emboldened Min Aung Hlaing to launch the coup because it meant the military was no longer engaged in a major conflict anywhere in Myanmar.Hide Footnote  Typically, the military would try to disrupt any attempt by an armed group to weaken state control and establish parallel administrative structures; it has done so for decades in various parts of the country where ethnic armed groups operate and has over the past year used brute force to crush similar “people’s administrations” set up by anti-coup forces.[fn]“Leading from the shadows: The successes and failures of the CRPH”, Frontier Myanmar, 20 March 2021.Hide Footnote  But more than fifteen months after the coup, the military remains locked in a potentially existential struggle elsewhere in the country with Burman and ethnic minority armed resistance forces that are increasingly coordinating on political and military affairs.[fn]Horsey, “One Year On from the Myanmar Coup”, op. cit.Hide Footnote  At a time when the regime’s forces are already stretched, it can ill afford to open another front – least of all with an opponent like the Arakan Army, which ranks among the country’s most powerful armed groups. It has thus largely refrained from confronting the Arakan Army militarily, making it much easier for the group to consolidate control.[fn]A similar phenomenon has unfolded in northern Shan State, where the military has not intervened to stop the Ta’ang National Liberation Army (TNLA) and Shan State Progress Party from seizing territory from a rival armed group, the Restoration Council of Shan State, that is considered closer to the regime. Having dislodged the third group, the TNLA is presently building an administrative system similar to that of the Arakan Army. See “Rising dragon: TNLA declares ‘victory’ in northern Shan”, Frontier Myanmar, 4 February 2022, and “Wa an early winner of Myanmar’s post-coup war”, Asia Times, 22 February 2022.Hide Footnote

The military has at times gone even further, actively seeking to maintain cordial relations with the Rakhine armed group. It has, for example, taken the Arakan Army off its list of terrorist organisations, freed scores of people arrested for alleged links to the group, invited its representatives to peace talks alongside other ethnic armed groups, welcomed its representatives to Union Day celebrations in Naypyitaw, and even cooperated in distribution of COVID-19 vaccines in rebel-controlled areas.[fn]See, for example, “China facilitates Myanmar junta and ethnic armies’ talks”, The Irrawaddy, 16 December 2021; “Why the Arakan Army attended Myanmar junta’s Union Day event”, The Irrawaddy, 16 February 2022; and “Northern Arakan targeted in vaccine push jointly undertaken by AA and military govt”, Development Media Group, 25 January 2022.Hide Footnote  On the few occasions when there have been confrontations between military and Arakan Army forces on the ground – such as the clashes that erupted in November 2021 and February 2022 in northern Maungdaw Township – local commanders from both sides moved quickly to de-escalate tensions.[fn]“Rumbles in Rakhine amid strains between Myanmar military, rebels”, Al Jazeera, 24 November 2021; “Junta troops clash with Arakan Army in western Myanmar”, The Irrawaddy, 8 February 2022; and “Military, Arakan Army confront each other near Kyauktaw”, Myanmar Now, 19 January 2022.Hide Footnote

“Rumbles in Rakhine amid strains between Myanmar military, rebels”, Al Jazeera, 24 November 2021; “Junta troops clash with Arakan Army in western Myanmar”, The Irrawaddy, 8 February 2022; and “Military, Arakan Army confront each other near Kyauktaw”, Myanmar Now, 19 January 2022.

Their alacrity reflects a shared desire to avoid a return to war, at least for now.

The Arakan Army’s task has been made much easier by the composition of the civil service in Rakhine State, where the vast majority of civil servants are local, including in some of the highest positions of the state’s bureaucracy.[fn]Crisis Group interviews, various Rakhine-based sources, January 2022.Hide Footnote  In most other ethnic minority states, by contrast, ethnic Burmans deployed from the country’s Bamar heartland make up a much higher proportion of the civil service.[fn]Ibid.Hide Footnote  Due to ethnic solidarity, many civil servants in Rakhine have thus been willing to cooperate with, if not actively support, the Arakan Army. A teacher observed: “I don’t think the Arakan Army needs to take over the running of schools – because over 90 per cent of teachers in Rakhine State are ethnic Rakhine, they are basically under the Arakan Army’s influence already”.[fn]Crisis Group interview, state schoolteacher, January 2022.Hide Footnote

Even those unsympathetic to the Arakan Army cause are … willing to cooperate with the group’s governance structures.

Since the fighting ended in November 2020, some civil servants have resigned to join the Arakan People’s Authority, including members of the judiciary, as mentioned above, but also police officers and administrators. Many others work for the group while continuing to nominally serve Naypyitaw. Even those unsympathetic to the Arakan Army’s cause are in many cases willing to cooperate with the group’s governance structures. “On the ground, many [regime-employed] officials cooperate with the ULA political and governance system. Only those from the military side and high-ranking civil servants resist”, said one Rakhine-based political analyst.[fn]Crisis Group interview, Rakhine-based political analyst, January 2022.Hide Footnote  Several interviewees also noted that it was common for state-employed civil servants to pay taxes to the Arakan Army.[fn]Crisis Group interviews, various Rakhine-based sources, January 2022.Hide Footnote

The military regime has not given the Arakan Army an entirely free hand, however. Particularly since late 2021, it has increased troop deployments in some strategically important areas of the state – notably in Maungdaw Township on the border with Bangladesh and Kyauktaw township, which is the main gateway to Paletwa in neighbouring Chin State. It has also pressured Rakhine and Rohingya community leaders to avoid interacting with the Arakan Army and Arakan People’s Authority (see section V below) and set up new checkpoints in several townships in an apparent effort to prevent the insurgents from moving forces around the state.[fn]See, for example, “Military tightens security checks on travellers in some Arakan townships”, Development Media Group, 4 January 2022.Hide Footnote  It has also detained young Rakhine suspected of training with the Arakan Army, as well as others thought to be providing support. [fn]See “Dozens arrested at northern Maungdaw security checkpoint”, Development Media Group, 31 January 2022; and “Regime troops’ visits to Arakan State villages are increasingly common, and unsettling, for locals”, Development Media Group, 9 May 2022.Hide Footnote  The Arakan Army does not seem deterred by these measures, as underscored by Twan Mrat Naing’s expletive-filled May 2022 tweet warning the local military commander not to go “too far”.[fn]See tweet by Twan Mrat Naing, @TwanMrat, Arakan Army leader, 5:25am, 6 May 2021.Hide Footnote  All this evidence points to rising tensions in Rakhine State in recent months and raises the risk of a return to war (see Section VI.A below).

IV. The Rohingya: Caught in the Middle

The rise of the Arakan Army has also had an impact on the fate of Rakhine state’s Rohingya Muslim population, now estimated at approximately 620,000 after over 700,000 fled to Bangladesh in 2016-2017.[fn]“Humanitarian Response Plan Myanmar”, UN Office for the Coordination of Humanitarian Affairs, January 2022.Hide Footnote  On paper, central government regulations ban Rohingya from travelling freely outside the township in which they reside and significantly limit their access to public services, including education and health care. But given that most Rohingya live in the state’s north, a significant number are now in areas largely under Arakan Army control, particularly in Buthidaung and Kyauktaw townships. Many also live in contested areas where the group exerts significant influence but not outright control, such as northern Maungdaw and Pauktaw townships. Rohingya Crisis Group spoke to report mixed outcomes from the erosion of state control: many consider it an improvement, particularly in terms of freedom of movement and access to Arakan People’s Authority services, but it has also come at a cost, as it has resulted in threats and new restrictions from the military.

Even prior to the Arakan Army asserting control in Rakhine State, its rise opened the way for an amelioration in relations with the Rakhine Buddhist majority. In 2012, communal conflict between Rakhine and Rohingya had resulted in around 200 deaths and 140,000 people – mostly Rohingya – being displaced, leaving the state’s Buddhists and Muslims deeply divided, often to the point of segregation.[fn]Crisis Group Asia Report N°261, Myanmar: The Politics of Rakhine State, 22 October 2014, Section III.B.Hide Footnote  Since then, neither local Rakhine politicians nor Naypyitaw appeared to have the political will to mend the damage done to communal relations. Rakhine civilians were implicated in attacks on Rohingya during the military’s 2017 campaign of targeted violence. The International Court of Justice is presently hearing claims that this campaign violated Myanmar’s obligations under the Genocide Convention.[fn]Notably, the National League for Democracy government failed to follow key recommendations made by the Kofi Annan-led Advisory Commission on Rakhine State. On the ICJ case, see Richard Horsey, “Myanmar at the International Court of Justice”, Crisis Group Commentary, 10 December 2019. On Rakhine militias, see Crisis Group Report, Resisting the Resistance, op. cit.Hide Footnote

The Arakan Army leadership has, by contrast, sought to build more positive relations with the Rohingya in recent years. In particular, its leadership has reframed the Rakhine struggle as a fight with the Burmans, Myanmar’s largest ethnic group, and explicitly said the state’s Rohingya population should not be seen as the enemy. Arakan Army figures have also articulated a “nation-building” agenda that includes the creation of a more tolerant, inclusive “Arakan” identity that encompasses all groups living in the state, including Muslims.[fn]“Interview with ULA chairman and AA commander-in-chief Twan Mrat Naing”, op. cit.Hide Footnote  Although they have stopped short of officially endorsing the term “Rohingya”, which is highly contested in Myanmar, the top leader, Twan Mrat Naing, has used it in interviews.[fn]See “‘We recognise the human rights and citizen rights of the Rohingyas’”, Prothom Alo English, 3 January 2022.Hide Footnote

The outbreak of war between the Arakan Army and military in late 2018 reinforced a narrative that cast Burmans as the enemy, and Rakhine anger at the Rohingya almost immediately shifted to civilian and military leaders in Naypyitaw. Numerous sources noted that relations between Rohingya and Rakhine have improved since then, although interactions are still often limited to economic activities.[fn]Crisis Group interviews, various sources based in Rakhine State, January-March 2022.Hide Footnote

The Arakan Army has not always appeared so sympathetic to the Rohingya.

The Arakan Army has not always appeared so sympathetic to the Rohingya. After the Arakan Rohingya Salvation Army (ARSA) attacked police outposts in October 2016, the Arakan Army described the group as “savage Bengali Muslim terrorists”; Bengali is considered offensive by many Rohingya, as it implies they are immigrants from Bangladesh.[fn]ULA/AA press release, 20 October 2016, as cited in David Scott Mathieson, “The Arakan Army in Myanmar: Deadly Conflict Rises in Rakhine State”, U.S. Institute of Peace, November 2020.Hide Footnote  The following year, Twan Mrat Naing used the derogatory term kalar when he described the Rohingya “problem” as a “political trap” for the Rakhine.[fn]“AA chief urges Arakanese not to fall into army trap in Rakhine”, The Irrawaddy, 11 December 2017. The word kalar was originally used to describe all foreigners who had come to Myanmar (then Burma) by sea, but in modern times it is used primarily to denote South Asians, particularly Muslims, and is widely considered racist. For a further discussion of the etymology, see Matt Schissler, Matthew J. Walton and Phyu Phyu Thi, “Reconciling Contradictions: Buddhist-Muslim Violence, Narrative Making and Memory in Myanmar”, Journal of Contemporary Asia, vol. 47, no. 3 (2017). In 2020, activists launched a campaign, “Don’t call me kalar”, to highlight its racist connotations. See “Challenging entrenched racism in Myanmar: Don’t call me ‘kalar’”, Progressive Voice, 18 June 2020.Hide Footnote  But by August 2020, Twan Mrat Naing was tweeting “Happy Eid Mubarak to you all!”, a holiday greeting to Muslims that would have been unimaginable a few years earlier.[fn]Tweet by Twan Mrat Naing, @TwanMrat, Arakan Army leader, 5:54pm, 2 August 2020.Hide Footnote  Similarly, the group issued condolences following the death of Wakar Uddin, a prominent Rohingya leader based in the U.S. in April 2022.[fn]“Letter of Condolences”, United League of Arakan, 1 May 2022.Hide Footnote

This seemingly pragmatic evolution reflects the different stages of the Arakan Army’s struggle. In earlier phases, the group pitched a more exclusionary Rakhine ethno-nationalism to the Rakhine people as it sought to build support for its cause. After fighting with the military erupted in December 2018, it became increasingly concerned about its international image, and sought to present itself as more tolerant of the Rohingya than the Myanmar state and the Tatmadaw. During the conflict, it also made sense to have cordial, if not positive, relations with the Rohingya.[fn]For a closer examination of the Arakan Army’s position on the Rohingya, see Mathieson, op. cit., p. 16.Hide Footnote  Fighting a guerrilla war, the rebels depended on their capacity to conceal their forces and to survive on support from residents of the areas where they were operating. It therefore helped not to have the Rohingya as an enemy.

The armed group’s relationship with the Rohingya became even more important after it began administering territory in which Rohingya lived. As it has rolled out administrative structures, the group has worked to include Rohingya leaders in local administration, including on committees that resolve disputes. Rohingya administrators told Crisis Group they preferred working with these new structures to engaging with the military regime. As one noted:

Because the Arakan Army is working for all communities, we prefer their system, and we work together. We ignore the orders from the junta.[fn]Crisis Group interview, Rohingya administrator from Kyauktaw, January 2022.Hide Footnote

Although these comments reflect a degree of genuine improvement in living conditions for Rohingya in areas under Arakan Army control, such as reduced restrictions on movement, they should be understood in context. The expression of positive sentiments about the Arakan Army needs to be considered in light of the extreme repression that Naypyitaw has inflicted on the Rohingya for decades, which makes even modest betterment of the situation seem quite positive. Rohingya may also not feel comfortable openly criticising the Arakan Army, given that they are living in areas under its control. Moreover, reports are not uniformly positive. Some sources suggest that Rohingya experiences with Arakan People’s Authority structures, such as its judicial system, have been uneven, with complaints of bribes being requested in some areas.[fn]Crisis Group interviews, various sources in Rakhine State, January-March 2022.Hide Footnote

Some Rohingya may also perceive the new parallel governance structures as an extra burden, forcing them to pay taxes to the armed group while they still need to interact with state bureaucrats – and pay bribes – for example to obtain permits to travel to the state capital, Sittwe, or other military-controlled areas.[fn]Crisis Group interview, January 2022.Hide Footnote  Another complaint about the Arakan People’s Authority is the lack of inclusion of Rohingya and other ethnic minorities in its middle and upper levels.[fn]Crisis Group interviews, various sources in Rakhine State, January-March 2022.Hide Footnote

Beginning in August 2021, military officials based in Rakhine began warning Rohingya community leaders not to have any contact with the Arakan Army or the Arakan People’s Authority.[fn]Crisis Group interviews, various Rohingya sources in Rakhine State, January 2022: and local media reports.Hide Footnote  At one meeting in Kyauktaw Township, officers claimed that the Rohingya could rely on the state’s administration and judicial systems. But a Rohingya community leader present at the meeting told Crisis Group there was little choice but to use the Arakan People’s Authority, because state structures no longer exist in the area. “Officially I work for the [regime], but I haven’t seen any officials in our area for at least two years. Even since the coup, junta members have been too scared to come to our village”, he said, adding: “The Arakan Army is good for us. If we have problems, they come and solve them”.[fn]Crisis Group interview, Rohingya administrator from Kyauktaw, January 2022.Hide Footnote

Beyond repeated warnings to community leaders, the military has imposed increased restrictions on the Rohingya. In November 2021, it tightened the process of issuing travel permits for movement beyond township limits in Maungdaw and Buthidaung, where the vast majority of Rohingya live. Community members in these areas now need to get permits from local Immigration Department branches, in addition to the longstanding requirement that they obtain approval from local state administrators.[fn]Crisis Group interview, Rohingya community leader from Buthidaung, January 2022.Hide Footnote

The modest improvements some Rohingya have seen in territory controlled by the Arakan Army should not obscure the economic plight that the community faces.

Finally, the modest improvements some Rohingya have seen in territory controlled by the Arakan Army should not obscure the economic plight that the community faces throughout Rakhine State. The two-year war in Rakhine, the COVID-19 pandemic and Myanmar’s post-coup economic crisis have created further financial hardship for the Muslim minority, which was already among the most impoverished groups in Myanmar. “Since the coup, we don’t have any work – we are just trying to survive, trying not to die”, said the Rohingya community leader from Buthidaung.[fn]Crisis Group interview, Rohingya community leader from Buthidaung, January 2022.Hide Footnote

Combined with the political uncertainty the community faces, this hardship is prompting many to try to leave the state, often in the hope of reaching Malaysia, which already has a large Rohingya community. This illegal journey is expensive and risky; police regularly detain large groups of Rohingya in other parts of the country and charge them with travelling outside their home township without permission, an offence usually punished with two years’ imprisonment.[fn]Several hundred Rohingya have already been arrested in 2022 in other parts of Myanmar, nearly all of them while attempting to reach a third country (usually Malaysia). In one notable example, regime police arrested 65 Rohingya in the town of Myawaddy, on the border with Thailand, as they sought to reach Malaysia. See “65 Muslims arrested in Myawaddy for illegally travelling to Malaysia”, Narinjara, 3 March 2022.Hide Footnote  Those who try to escape by sea face a perilous voyage at the hands of human traffickers.

V. A Clouded Future

A. A Return to War?

Despite the calm in Rakhine State since the November 2020 informal ceasefire and the subsequent coup, a return to conflict remains possible, even likely. The lack of a formal ceasefire, particularly one with clearly demarcated lines to separate the forces, means that troop movements can easily result in clashes in disputed areas, as was the case with fighting in Maungdaw in late 2021 and early 2022.

The Arakan Army’s objective of confederacy, if not independence, is anathema to the military, which envisions a centralised, unitary state, and has sought to maintain control of the country’s border areas. Although the military has largely tolerated the Arakan Army’s state-building agenda since the coup, the scale of the undertaking is beginning to provoke stronger pushback from the regime, particularly as the Arakan Army seeks to expand into areas that the military deems strategic. Interviews with Rakhine and Rohingya residents and other sources in the state indicate that many believe a return to war is only a matter of time.[fn]Crisis Group interviews, various sources in Rakhine State, January-March 2022.Hide Footnote  After first seeking to downplay its administrative expansion, the Arakan Army has grown increasingly bold by attempting to expand into the Bangladesh border region, talking more publicly about its parallel administrative system and openly describing itself as a “government”.[fn]The Arakan Army has described itself as a “Rakhine transitional government” to aid workers and at the 5 March press conference Khaing Thu Kha referred to it as a pyithu asoya, or “people’s government”.Hide Footnote

A renewed outbreak of conflict could be even more devastating than the two-year war that ravaged parts of Rakhine State in 2018-2020. The Arakan Army has significantly built up its armed forces since the ceasefire, with as many as 30,000 people having now undergone training.[fn]Crisis Group interview, source close to the military regime, March 2022. See also “Rebel yell”, op. cit.Hide Footnote  Newer members lack combat experience, and possibly also arms and equipment, but the insurgency will nevertheless be an even more formidable opponent for the regime than it was during the earlier conflict, when it inflicted significant casualties on the military. The strength of popular support it enjoys will enable its fighters to remain highly mobile; they will also benefit from Rakhine communities providing intelligence and supplies, thus making a Tatmadaw victory close to impossible. Still, the humanitarian consequences of renewed fighting would likely be devastating for both Rakhine and Rohingya residents, leading to large-scale displacement.

War is not inevitable, however, at least not in the near term. All will depend on the political will of the junta and Arakan Army leadership to find a mutually acceptable near-term arrangement that moves beyond the fragile status quo. A formal ceasefire agreement that demarcates territory in Rakhine State could avert a return to conflict and create greater stability for local populations. It would preserve both parties’ current positions, allowing for the possibility of more substantive negotiations on an Arakan Army-controlled autonomous region to take place at a later date.

The regime has not only been forced to accept the fact of the Arakan Army’s presence, but ... appears to be open to an agreement that would formally recognise it.

A formal ceasefire agreement had seemed highly unlikely before the coup: Myanmar’s military leadership had demanded that the Arakan Army withdraw from Rakhine State and return to Kachin State, where the group was first established.[fn]Crisis Group Report, An Avoidable War, op. cit.Hide Footnote  The informal ceasefire, though, reflected a shift in the Tatmadaw’s position and this has only solidified amid post-coup realities. The regime has not only been forced to accept the fact of the Arakan Army’s presence, but according to several sources Crisis Group spoke to, appears to be open to an agreement that would formally recognise it.[fn]Crisis Group interviews, sources with knowledge of informal discussions between the military and the Arakan Army, January-April 2022.Hide Footnote  Although meetings between the two sides are rare, both mid-level military commanders and political intermediaries are in regular contact.[fn]Crisis Group interviews, sources close to the military regime, March and April 2022.Hide Footnote  Such a shift in approach from the military means a formal ceasefire is now possible.

To place these developments in context, late last year, junta leader Min Aung Hlaing also launched a new peace initiative and has declared 2022 a “year of peace”. His primary motive is to undermine opposition to the coup by enticing ethnic armed groups from the battlefield to the negotiating table, enabling the military to focus its forces on newly formed resistance groups, many of which are aligned with the NUG. He also hopes the initiative will distract outside actors and stave off pressure to open talks with the NUG and other resistance groups. The process lacks legitimacy and is highly unlikely to lead to a comprehensive settlement that ends Myanmar’s internal conflicts. But Min Aung Hlaing’s eagerness to cut deals and the junta’s overall weak position create an opportunity for the Arakan Army to use this initiative to negotiate an arrangement that might otherwise be unattainable.

Past experience in Myanmar suggests that the parties could reach such an agreement quickly. With the exception of the Nationwide Ceasefire Agreement, most ceasefires are light on detail – if they are written down at all – and are instead based on understandings or “gentlemen’s agreements” between military officials and armed group leaders. In some cases, these understandings have endured for decades.

Still, the two sides are likely to be some distance apart on the key terms of such an arrangement. In particular, the military may not accept the present geographic extent of Arakan Army control or meet the group’s demands on the level of autonomy it wishes to enjoy. There are also other reasons for both sides to be wary; in particular, a ceasefire without an explicit political element would enable their opponents to build up their forces and prepare for renewed confrontation down the road.

Nevertheless, both sides have compelling reasons to try reaching a deal. The regime’s new peace process has so far had limited buy-in from the country’s ethnic armed groups; if the Arakan Army were to engage in ceasefire negotiations through this process, it would give the process a major boost that the military would likely welcome.[fn]The military invited all ethnic armed groups – but not the National Unity Government or its People’s Defence Forces – to a ceremony to mark Union Day on 12 February in Naypyitaw, but only ten of Myanmar’s twenty or so ethnic armies sent delegations. One of them was the Arakan Army.Hide Footnote  A ceasefire would also deprive the military’s opponents, particularly the NUG, of a potentially powerful ally, at least for now. For the insurgents, a formal ceasefire would be an opportunity to further consolidate control over their newly won territory. Moreover, any agreement that explicitly or implicitly recognises their authority would represent a major political victory, even if at first it does not deliver the level of autonomy that they seek. It would also enable other actors, such as humanitarian aid groups and Bangladesh, to more openly engage with the Arakan Army, something the armed group has been seeking.

While the coup has put a previously unattainable deal within reach, it also complicates negotiations.

But while the coup has put a previously unattainable deal within reach, it also complicates negotiations. The Arakan Army leadership has already been forced to walk a fine line between engaging the junta in pursuit of its own political agenda and avoiding the appearance that it is legitimising the regime, which remains extremely unpopular in Rakhine State and Myanmar more broadly. At the 5 March press conference, Khaing Thu Kha asserted that the Arakan Army does not recognise the junta as the legitimate government and that it was “impossible” to hold a “political dialogue” with an administration that had seized power in a coup. Any political agreement, for example on autonomy and related issues, may therefore have to wait, at least until after a planned 2023 election. But in the interim, both sides could hold negotiations on a ceasefire focused on military matters, particularly demarcation of territory.

Pending the result of such negotiations, both sides should avoid taking steps that might provoke a return to fighting. The red lines for the military are blurry, but they are likely to include any attempt by the Arakan Army to gain control of the northern border areas with Bangladesh, townships in southern Rakhine (including Ann) or urban centres in the centre and north of the state. Optics are also likely to be important for the junta, so the group should refrain from showcasing the extent of its military and administrative control, so as not to portray itself as a de facto government. The military, meanwhile, will need to accept the fact of Arakan Army control in most rural areas of the state. It should stop aggressive troop movements and its attempts to disrupt the activities of the Arakan People’s Authority through arrests and other forms of pressure on civilians.

One alternative for the Arakan Army is to increase political and military cooperation with the NUG, as some other ethnic armed groups have done since the coup. The Arakan Army has held numerous informal meetings with the NUG, which hopes to entice the pro-Rakhine group to join the anti-military resistance. Many in Myanmar see the Arakan Army as a potential game-changer in the battle with the military regime. But the NUG has been unwilling to meet the Arakan Army’s demands for autonomy, which goes beyond the federal system that the rebel administration is negotiating with ethnic armed groups and other stakeholders.[fn]Crisis Group interview, NUG sources, March 2022.Hide Footnote  There is also still deep antagonism among many Rakhine, including the Arakan Army leadership, toward the National League for Democracy (NLD) for supporting the military’s war against the Arakan Army in 2018-2020.[fn]Crisis Group Report, An Avoidable War, op. cit.Hide Footnote  By joining with the NUG, the Arakan Army risks provoking a return to war with the junta, something it still seems keen to avoid, and which would exact a heavy toll on the people of Rakhine State.

There are other potential pathways forward as well. The Arakan Army might, for example, pursue a formal ceasefire with the military while also continuing to engage the NUG, as well as other resistance forces and ethnic armed groups. The group might see strategic benefit in keeping its options open, given the uncertainty about how Myanmar’s broader conflicts will play out. If the Arakan Army tries to go this route, however, it could make the military regime less amenable to its ceasefire demands, particularly if the latter continues to publicise its talks with the NUG and support for resistance groups. Conversely, the decision to pursue negotiations within Min Aung Hlaing’s peace initiative or separate from it would likely increase the military’s willingness to meet the Arakan Army’s terms, as it will clearly prefer that the armed group join the talks, thus bolstering their credibility.

B. Implications for Aid Delivery

Rakhine State has historically been a major destination for humanitarian aid in Myanmar, due to its high overall poverty and the needs of its Rohingya population in light of discriminatory policies. But humanitarian activities have been impeded by post-2012 communal tensions, the military’s violent campaign against the Rohingya in 2016-2017, the war between the Arakan Army and the military between 2018 and 2020, and the outbreak of COVID-19.

After taking power, the military regime relaxed some restrictions, likely in an attempt to improve its international image vis-à-vis the previous NLD government.[fn]Crisis Group interview, aid agency official, March 2022. Close to 150,000 Rohingya displaced by the 2012 communal violence remain in camps, villages and other displacement sites, almost entirely dependent on international humanitarian aid. “Number of internally displaced in Myanmar doubles, to 800,000”, UN News, 11 February 2022.Hide Footnote Despite these limited improvements, however, at the time of writing, aid workers estimated that around 75 per cent of humanitarian activities in the state were still facing constraints, including insecurity, blanket bans on access to certain areas, denials or delays in approving travel authorisations, conditions or limitations on the authorisations that are issued, or demands from the authorities for detailed reports.[fn]Crisis Group interviews, humanitarian workers and aid agency officials, March 2022.Hide Footnote  Some organisations have scaled back operations or even closed entirely due to these restrictions.[fn]Crisis Group interview, aid agency official, March 2022.Hide Footnote

In general, the Arakan Army’s increased assertion of administrative control has not significantly affected humanitarian operations. Aid workers have been able to pass through rebel checkpoints by showing Naypyitaw-issued travel authorisations and the armed group has not otherwise sought to stop them from carrying out their activities.[fn]Ibid.Hide Footnote  Local civil society organisations, which are not subject to the travel authorisation system, similarly report few restrictions. “Arakan Army officials just asked us not to take too many photos while we are in their areas. They said this is for security reasons – they don’t want outsiders to get too much information about what they are doing”, said one civil society leader, whose organisation works mostly in areas under the group’s control.[fn]Crisis Group interview, civil society leader in Rakhine State, January 2022.Hide Footnote  Another civil society leader from Ramree said that, when his staff informs regime officials about their planned activities, they are usually given a long list of restrictions, whereas the Arakan Army always gives permission. “We prefer to work with the Arakan Army’s administration. … The [regime] restricts and prohibits many things”.[fn]Crisis Group interview, civil society leader in Ramree Township, January 2022.Hide Footnote

Recently, however, the Arakan Army has begun exerting greater control over aid activities within the territory it controls, particularly those implemented or funded by international organisations. Foreign aid workers report that the armed group has been contacting local implementing partners to seek information about their activities and personnel, and in some cases asking them to register with its parallel administration.[fn]Crisis Group interview, aid agency official, March 2022.Hide Footnote  The group has set up a dedicated Humanitarian and Development Coordination Office to engage with the UN and international NGOs but has not elaborated on how this office will work in practice.[fn]Ibid.Hide Footnote  The office reflects both the insurgency’s desire to be treated as a state-like entity and the significant role that humanitarian actors play in Rakhine State. The Arakan Army has claimed the plan would ease operational difficulties they are facing, but some fear that introducing such reporting requirements could further constrict the humanitarian space in Rakhine State by forcing aid agencies to report to two parallel administrations.[fn]Crisis Group interviews, humanitarian workers and aid agency officials, March 2022. Aside from adding an extra layer of bureaucracy that restricts humanitarian activities and consumes resources, a parallel system would be impractical unless travel authorisations issued by the regime and the Arakan Army were near identical. Travel authorisations are for a specific time period, for example, and the time period would need to align on both authorisations for activities to be implemented.Hide Footnote

The Arakan Army should ... tread carefully as it devises a system to interact with aid organisations.

The Arakan Army should therefore tread carefully as it devises a system to interact with aid organisations, avoiding the introduction of new rules that in practice restrict their capacity to operate, and respecting the need for these organisations to work with Naypyitaw, especially as many of them have operations in other parts of the country. In particular, it should not force them to register, pay taxes or obtain travel authorisations from a parallel system it would introduce. Unworkable rules that force aid groups to scale back support to the people of Rakhine State would harm its own interests – negatively affecting how the Arakan Army is perceived by both residents of Rakhine State, who would be deprived of aid, and by foreign governments, many of whom have provided humanitarian funding to Rakhine for years.

For their part, international organisations working in Rakhine, including UN agencies, should forge a common approach to how and in what circumstances to engage the Arakan Army. Because Rakhine has certain unique features relative to other parts of Myanmar where ethnic armed groups are also present, there is no template to borrow from; this framework will have to be developed more or less from scratch.[fn]There are key differences in both scale and timing. Although humanitarian organisations are active to some degree in other insurgent-controlled areas of Myanmar, they usually operate only through local implementing partners and on a much smaller scale than in Rakhine. Further, because most of these other insurgencies date back decades, aid agencies were generally not present prior to an armed group establishing control, as has been the case in Rakhine. A further difference is that humanitarian aid has been at times a highly divisive issue in Rakhine State, with many ethnic Rakhine feeling that the Rohingya have received preferential treatment from aid actors.Hide Footnote  In addition to bringing them together to work on this framework, such a coordinated approach will give aid organisations maximum leverage if they need to push back against a parallel system of authorisations. Engagement with the Arakan Army should not be seen as just a risk, however; it could also be an opportunity to positively influence the group’s policies and practices.

C. Rohingya Repatriation

The rise of the Arakan Army in Rakhine State will inevitably affect prospects for the repatriation of Rohingya refugees, particularly over the long term. An estimated 730,000 Rohingya fled to Bangladesh in 2016-2017 due to the Myanmar military’s brutal crackdown on the community in northern Rakhine State. Bangladesh and Myanmar signed a bilateral repatriation agreement in November 2017, but no refugees have yet returned through formal channels despite two high-profile attempts by Bangladesh in November 2018 and August 2019.[fn]For a detailed examination of the issues around repatriation prior to the military coup, see Crisis Group Asia Briefing N°153, Bangladesh-Myanmar: The Danger of Force Rohingya Repatriation, 12 November 2018.Hide Footnote

Refugees have cited the failure of the Myanmar authorities to guarantee their safety in Rakhine State and the inability to get citizenship as the main reasons for refusing to return. At the time, the NLD government largely ignored these demands and instead blamed Bangladesh for bureaucratic delays that it said were holding up repatriation.[fn]“Myanmar blames Bangladesh for Rohingya repatriation failure”, The Irrawaddy, 18 November 2019.Hide Footnote  Myanmar has also said ARSA militants in the camps that house hundreds of thousands of Rohingya in southern Bangladesh have been threatening refugees into rejecting repatriation; Bangladesh has acknowledged that some refugees are using force to oppose repatriation, but until very recently it rejected the suggestion ARSA is active in the camps.[fn]See, for example, “Bangladesh’s harboring of terrorists continues to hinder Rohingya repatriation process”, The Irrawaddy, 13 October 2020. Bangladesh maintained for years that ARSA was not present in the camps and that alleged ARSA members there were actually criminals using the group’s name. After the killing of Mohib Ullah, a Rohingya political activist, in September 2021, Dhaka has acknowledged ARSA’s presence in the camps. See “Armed group behind Rohingya leader’s murder: Bangladesh police”, Al Jazeera, 16 March 2022.Hide Footnote  Despite China’s efforts to mediate between Dhaka and Naypyitaw, the repatriation process stalled entirely after large-scale fighting erupted in Rakhine state in 2019.

The coup has only made the situation more challenging. It dealt a further blow to prospects of large-scale repatriation, as many refugees will likely remain reluctant to return with the military that was directly responsible for the mass violence that drove them from the country in power.

The coup also delayed the resumption of bilateral negotiations with Bangladesh on repatriation, but it has not derailed the process entirely. Dhaka has been understandably anxious to start sending refugees back – or at least avoid the process being derailed entirely – and has ensured repatriation remains the central issue in bilateral relations. For its part, the Myanmar military has stated since shortly after the coup that it will continue repatriation efforts. Although this is almost certainly an attempt to improve its international image, it leaves the door open for further discussions.[fn]In his first major public address after the coup, Min Aung Hlaing said his administration would “continue receiving the displaced persons in Bangladesh in accord with the bilateral agreement”. Since then, the regime has blamed Bangladesh for further delays in repatriations. See “Republic of the Union of Myanmar State Administration Council Chairman Senior General Min Aung Hlaing makes speech to public”, Global New Light of Myanmar, 9 February 2021; “MOFA Press Release on ASEAN Foreign Ministers’ Meeting Retreat”, Myanmar Ministry of Foreign Affairs, 20 February 2022; and “Rohingya refugees reject return to Myanmar without assurances”, Radio Free Asia, 24 February 2022.Hide Footnote  In January 2022, Myanmar and Bangladesh resumed talks via a virtual coordination meeting and in March the Rakhine State Administration Council said it had received a list of what it referred to as 700 “Muslims” to be repatriated, though no further details have since been released by either country.[fn]See “Bangladesh, Myanmar resume talks over Rohingya verification”, The Independent, 28 January 2022; and “ဘင်္ဂလားဒေ့ရှ်မှ မူဆလင် ၇ဝဝ ကျော်ကို ပြန်လည်လက်ခံရန်ရှိဟု ရခိုင်ပြည်နယ်ကောင်စီပြော [More than 700 Muslims who fled to Bangladesh to be repatriated, says Rakhine State Administration Council]”, Narinjara, 8 March 2022.Hide Footnote

[The return of some Rohingya refugees to Rakhine State] mostly reflects the deterioration of conditions in the camps in southern Bangladesh.

Despite reservations about returning to military-controlled Myanmar without guarantees on citizenship and other issues, some Rohingya refugees may be willing to consider heading back to Rakhine State. Indeed, an unknown number – possibly in the low thousands – have already returned informally, crossing the border on their own. While their return may be partly due to the November 2020 de facto ceasefire, which brought some level of stability in Rakhine State, it mostly reflects the deterioration of conditions in the camps in southern Bangladesh, where reports of crime and violence have been steadily increasing.[fn]“Bangladesh concerned by growing crimes, unrest at Rohingya camps”, Dhaka Tribune, 2 October 2021.Hide Footnote  Dhaka has also been taking a tougher approach to the refugees, closing some schools, destroying shops in the camps.[fn]See “Bangladesh destroys 3,000 shops belonging to Rohingya Muslim refugees”, The Independent, 5 January 2022; “Bangladesh: Rohingya Refugee Schools Face Closure”, Human Rights Watch, 18 December 2021; and “Bangladesh shuts largest private school in Rohingya camps”, Agence France-Presse, 28 March 2022.Hide Footnote  It has also moved close to 28,000 Rohingya to the island settlement of Bhasan Char.[fn]The Bangladesh government has set a target of relocating 100,000 refugees to Bhasan Char.Hide Footnote  Any formal repatriation would, however, likely involve only a small number of refugees, at least at first, for the reasons noted above.[fn]“Rohingya refugees reject return to Myanmar without assurances”, op. cit.Hide Footnote

Given the extent of Arakan Army influence in Rakhine, its explicit or tacit approval will likely be required to enable any large-scale organised repatriation to proceed. In line with its efforts to appeal to the Rohingya and improve its image abroad, leader Twan Mrat Naing has insisted his group does not oppose repatriation, saying it is “only natural” that Rohingya refugees would want to return. But he has also cautioned that the situation is not yet stable enough to initiate such a process and that a “massive” number of returns could lead to “unrest”.[fn]“‘We recognize the human rights and citizen rights of the Rohingya’”, op. cit. and “Rebel yell”, op. cit.Hide Footnote  Agreeing to support repatriation will be politically fraught for the group, as it is likely to draw complaints from its harder-line Rakhine supporters, who still consider most Rohingya to be illegal immigrants. Nevertheless, it may be willing to support limited formal returns if it believes it will gain enough, principally in enhanced legitimacy and international recognition. International actors should encourage it in that direction.

The emergence of the Arakan Army as a governance actor raises important questions for Bangladesh. The Arakan Army has long sought to build relations with the Bangladesh government, but Dhaka has rebuffed the overtures, because it has a policy of eschewing engagement with insurgencies that undermine the sovereignty of neighbouring states.[fn]“‘We recognize the human rights and citizen rights of the Rohingya’”, op. cit. On 21 February 2022, the United League of Arakan also issued a “statement of solidarity”, ostensibly to mark Bangladesh Language Martyrs Day, in which it spoke of the “natural bond between the people and regions of Bangladesh and Arakan”. It issued a similar statement to mark Bangladesh Independence Day the following month.Hide Footnote  With the Arakan Army and military regime no longer fighting and the armed group in partial or full control of much of the territory the Rohingya refugees on its soil originated from, Dhaka may wish to reconsider that approach in this particular case. Should the Arakan Army and Naypyitaw arrive at a formal ceasefire, Dhaka may reconsider it sooner.

VI. Conclusion

The two-year war and the subsequent de facto ceasefire between the Arakan Army and Myanmar military has radically reshaped the balance of power in Rakhine State. Taking advantage of the February 2021 coup, the armed group has eroded state control across much of Rakhine, and in its place rolled out its own governance and service delivery structures. Although these new mechanisms have faced some criticisms, support for the Arakan Army remains strong, particularly among the ethnic Rakhine. The Rohingya have also benefited from the group’s rise to some extent, but they remain vulnerable and wary of military retaliation for cooperating with the parallel administration.

Although under strain due to the nationwide uprising against the coup, the military is unlikely to tolerate an autonomous entity in Rakhine State. It may therefore attempt to dislodge the Arakan Army at some point. Another war would be devastating for everyone in Rakhine, and the Arakan Army is so well entrenched that Naypyitaw is unlikely to be able to wrest back power anyway. To avoid further conflict, both sides should reach a ceasefire agreement that recognises the armed group’s territory and increases stability in the state. Such an agreement would not only reduce the risk of further humanitarian disaster in Rakhine, but it would also bolster the prospects of return for at least some of the hundreds of thousands of Rohingya refugees presently lingering in refugee camps across the border.

Brussels, 1 June 2022

Appendix A: Map of Myanmar

Subscribe to Crisis Group’s Email Updates

Receive the best source of conflict analysis right in your inbox.