“Agents” of the state or revolution? Resistance, accommodation, violence and the role of local administrators in post-coup Myanmar

ABSTRACT

This paper examines the role of local authorities in Myanmar following a military coup in February 2021 which has generated a nation-wide resistance and put local state authorities in a difficult situation as mediators between the junta authorities, resistance movements, and local populations. Based on surveys of 26 community researchers, interviews of 51 residents in three most violent areas in Myanmar, and secondary sources published by state and non-state media, we find that local state administrators have adopted different responses and strategies depending on the types of authorities in control of a territory, and the level of local resistance to the military junta; the nature and scale of their loyalty to the military's regime; their relationship with the local residents; and the evolving political situation. These varying responses have in turn shaped whether they are perceived as “agents of the state,” or “agents of the revolution” by local populations and the resistance movement. The Myanmar case offers an insight on how the same type of actor (local administrators who are part of the state administrative machinery) can adopt two identities: agents of the state facilitating autocratization, or agents of the revolution resisting autocratization.

KEYWORDS:

• Local administrators
• resistance
• Myanmar
• military coup
• assassination

Introduction

Local authorities in war-torn countries are key to understanding how violence and a process of autocratization manifest themselves in specific sites. They occupy precarious and important roles as mediators between central authorities, resistance movements, and local populations. They can weaken the ability of central authorities to impose their will on society, or provide varying levels of assistance to resistance movements. These dynamics become even more acute under periods of civil war, where differing local governance arrangements emerge in state-controlled, hybrid, and rebel-controlled areas.

By examining the case of Myanmar, the article addresses the following questions: (1) What explains the different degrees of control that the military junta has over its local administrations; (2) how and why do different local administrators respond to similar types of threats against their lives? We find that variations in junta control over local administrations are explained by the types of authorities in control of a territory, and the level of local resistance to the military junta. The junta’s control over local state administration is stronger in areas that experience weak anti-coup resistance than in those with strong resistance. We also find that, while local state administrators from areas in the strongest resistance areas have increasingly used a low-profile approach to avoid threats against their survival, they have adopted different responses and strategies depending on the nature and scale of their loyalty to the military’s regime, their relationship with the local residents, and the evolving political situation. These varying responses have in turn shaped whether they are perceived as “agents of the state,” or “agents of the revolution” by local populations and the resistance movement. Those who have been targeted for assassination by the resistance movement tend to be portrayed by the local populations as “agents of the state” who provide crucial information about the resistance movement to the military, and/or those who enthusiastically carry out the military’s orders, and/or those who abuse their power and authority. In contrast, perceived “agents of the revolution” are those who half-heartedly implement military’s ordinances, while condoning the underground resistance movement and warning resistance groups of impending searches by security forces, or/and who remain politically neutral or are respected and trusted by both sides.

We focus particularly on the role of ward/village tract administrators (WA/VTAs).¹ They are part of the community, and yet they work as agents of the state and officially serve as the first point of contact with the government. They perform multiple functions including the maintenance of law and order, implementing government programmes and development projects, coordinating social and community activities, and mediating disputes.²

A case study on Myanmar contributes to existing studies by closing the gap between two disparate literatures on authoritarian governance and on civil war. The key insight from this article for the theme of this special issue is that agents of resistance have ambiguous identities and can emerge in a variety of ways in any regime types. In the Myanmar case, the same type of actor (local administrators who are part of the state administrative machinery) can adopt two identities: agents of the state facilitating autocratization, or agents of the revolution resisting autocratization.

Fragmented authoritarianism and the ambiguous roles of local authorities

A study of authoritarian governance and violence calls for a synthesis of two disparate bodies of literature on local governance in authoritarian but relatively politically stable countries on the one hand, and those in war-torn areas with fragmented authorities on the other. The former tends to focus on the ways in which central authorities use a hierarchical chain of command to implement policies through local authorities, while the latter tends to focus on how state and non-state actors compete for territorial and social control. There has not been a sufficient literature to fully illustrate and explain a wide range of local governance practices under various situations where state structures prevail, coexist, compete with, or are replaced by rebel governance structures. Our case study in Myanmar bridges this gap by explaining variations in military junta’s control over local state administrations across different environments, and in the responses by local authorities who are confronted with similar types of threats and dangers from competing armed groups.

Some studies on civil war show that state strategies towards rebel organizations and civilians in war-torn countries are influenced by a number of conditions. These include military capabilities and the degree of control each warring faction has over a particular region,³ geography and topography,⁴ the length of civil war, the extent of involvement by neighbouring countries, and the number of armed actors.⁵ On the other hand, rebel decisions on how to exert control over populations are influenced by the extent of government service provision,⁶ the availability of resources to finance rebellion,⁷ internal divisions within rebel leadership, and negotiation with aid organizations.⁸ Ana Arjona, for instance, categorizes different forms of rebel governance into “rebelocracy,” “disorder,” and “aliocracy,” which were determined by the time horizon of the group, and capability of civilians to act collectively and generate resistance.⁹ Kaplan also shows how rebel strategies to coopt, threaten, or intimidate local populations are influenced by the presence of state capacity, along with the strength of local authorities and civil societal organizations.¹⁰

Other literature on civil war examines responses by local communities towards competing authorities.¹¹ Kaplan shows how communities that are “caught between” the state and rebel groups resorted to various coping strategies to respond to multiple threats.¹² Marc Ashew’s article on southern Thailand’s violence-infested communities best captures the dilemmas of village heads and local populations in dealing with Thai state authorities and Islamic insurgent groups.¹³ Various literature shows how civilian decisions to align with the state or resistance, or to remain neutral, are influenced by the extent to which they are subject to state repression and exclusion, the type of local leadership and the role of external actors,¹⁴ pre-conflict organizational and kinship ties to the rebel movement,¹⁵ political and economic exclusion based on ethnic or identity categories,¹⁶ and low opportunity costs associated with rebellion.¹⁷ These situations in turn affect the strength and sustainability of a resistance movement.¹⁸

Literature on robust authoritarian countries focuses on the various circumstances under which local authorities have either strictly enforced central authorities or managed to evade or undermine central policies to protect local populations. In some cases, central authorities are able to effectively monitor the behaviour and practices of local authorities and enhance the resilience of authoritarianism at the local level.¹⁹ In other situations, local authorities align themselves with local populations to circumvent central policies to protect the interests of local communities.²⁰ Sometimes, local authorities further add their own layers of authoritarianism to abuse local populations.²¹ Several scholars highlight the circumstances under which local authorities are able to act as a “dual brokerage” between central authorities and local constituents.²² Jonathan Fox’s classic study on a “sandwich strategy,” on the other hand, focuses on the process of convergence between actions both from “above” (central governments) and “below” (local constituents) that can contain or weaken vested interests such as local authorities and businesses.²³

Yet, there has not been much literature that compares local administrations in areas controlled by the state, overlapping authorities, and by non-state actors. While Myanmar represents a class of countries often referred to as omnipresent and omnipotent authoritarian, it differs from many robust dictatorial countries in the extent to which its territories have never been under effective control of the state. A case study on war-torn Myanmar will close the gap between literature on authoritarian governance and civil war by providing a more nuanced and comparative perspective on state’s control over different local administrations across the country, and the various ways in which different local authorities under the same environment responded to the coup.

Research methods

This article is based on a survey filled out by 25 respondents across different parts of the country,²⁴ 51 in-depth interviews²⁵ in Yangon, Kayah, and Sagaing, two key informants’ interviews,²⁶ and analysis of secondary sources both on national aggregate data and on conventional and social media to analyse the narratives of the resistance movement.²⁷ A majority of 25 respondents who filled out the survey were local researchers who participated in a separate research project which monitors evolving political and economic circumstances in the post-military coup period. We used a snowball method to recruit our interviewees for in-depth interviews in Yangon, Sagaing, and Kayah (and four townships in Southern Shan adjacent to Kayah). These places are chosen for a preliminary exploratory study due to their accessibility (Figure 1; Table 1). Both interviews and surveys were collected in 2022.

Figure 1. All types of conflict in Myanmar (from 1 February 2021 to 22 Sept 2023). Source: Armed Conflict Location and Event Data (ACLED). Myanmar, 2021–2023. https://acleddata.com.

Table 1. Participants’ list.

	SAC-controlled areas with low resistance A	SAC-controlled areas with strong local resistance E	Mixed areas (armed groups friendly to SAC) with low resistance BF	Mixed areas (armed groups hostile to SAC) with low resistance C	Mixed areas with strong resistance G	Rebel-controlled areas DH
25 surveys^a	Naypyitaw, Northern Shan, Karen/Kayin, Ayeyarwady (1, 2), Mon (1), Rakhine (1,2), Kayah^b	Yangon^c, Sagaing (1)^d, Chin, Mandalay	Southern Shan 1,	Rakhine (3, 4), Kachin (1,2)	Bago, Magway Southern Shan 2^e, Rakhine (5, 6, 7) Tanintharyi	Sagaing (2)
51 interviews (Yangon, Kayah, Sagaing)^f 50 places plus one respondent describing general situation in Kayah	Yangon (total 13) Sagaing (total 7) Kayah (total 2)	Yangon (total 11) Sagaing (1)			Kayah (total 6) Southern Shan (total 4) Sagaing (total 6)

^a Described in states/regions.

^b The same respondent participated in one of the interviews in Kayah.

^c The same respondent participated in one of the interviews in Yangon.

^d The same respondent participated in the interview in Sagaing.

^e The same respondent participated in one of the interviews in Southern Shan.

^f Described in states/regions.

Transformation of the centres of violence in post-coup Myanmar

Despite Myanmar’s authoritarian history under military rule since 1962, political reforms by a quasi-democratic government led by the Union Solidarity Development Party (USDP), a military-affiliated political party, ultimately led to the electoral victory of the National League for Democracy (NLD), the main civilian opposition, in 2015. The NLD’s rule was characterized by increasing tensions with the military, which retained significant power on security matters under the 2008 constitution that it authored.²⁸ Such tension between the civilian government and the army eventually resulted in the military’s takeover in February 2021 after alleged voter fraud by the ruling NLD in elections the previous November.²⁹

The new regime, calling itself the State Administration Council (SAC), generated a nationwide resistance and unleashed an intense humanitarian crisis. It has also transformed the centre of conflict which used to concentrate mainly in ethnic minority areas.³⁰ They have hosted some of Myanmar’s oldest ethnic insurgent groups that have fought for greater political autonomy from the central government, and have been the site of a disproportionate share of violence and brutal behaviour by the Myanmar military. Prior to the coup, the majority Bamar ethnic group, who constitute about 65% of the population, were ignorant of or less sympathetic towards the plight of ethnic minorities, and Bamar-dominated regions served as traditional recruiting grounds for the Myanmar army. Data collected by ACLED shows that regions experiencing higher levels of violence and resistance are now Bamar-dominant regions which were historically NLD strongholds.³¹ The most violent place in Myanmar (Figure 1) is Sagaing region, measured in terms of fatalities, battles, violence against civilians, houses burnt. Mandalay and Yangon, two other Bamar-dominated regions, experience a low level of armed battles, but are ranked second and third in “violence against civilians.”

Most minority ethnic groups initially saw the coup as “Bamar fighting Bamar” and kept their distance from the conflict. Ethnic minorities have shared distrust and grievances against Bamar-dominated military and government including the Bamar-dominated NLD, which many believe used its popular legitimacy to marginalize minorities. The situation in ethnic minority communities, however, varies depending on the nature of the military’s relationships with ethnic armed groups in specific areas. Areas controlled by militias which have served as proxies of the Myanmar army, or those that have signed ceasefire agreements with the military, remained relatively peaceful. Some groups that engaged in war against the military in the pre-coup period have continued fighting against the army, and hosted and trained young protesters who fled the military’s crackdown.³² The areas they control now experience a higher level of armed conflicts and displaced populations.

While in reality the situations are more complex and fluid where local administrative practices have changed in response to the evolving situations, the above chart offers a simplified analytical framework which is confined to the period between Feburary 2021 and September 2023.³³

Our analytical framework in Figure 2 incorporates two sets of explanations for two different outcomes. We first show that variations in junta control over local administrations are explained by the types of authorities in control of a territory, and the level of local resistance to the junta. The junta’s control over local state administration is stronger in areas that experience weak anti-coup resistance than in those with strong resistance. In addition, the type of local administration affects administrator behaviour to the extent that local representatives in areas of strong state control over local administration are more likely to have a higher level of confidence in implementing junta policies, and the junta is able to replace local administrators who resigned from their posts or who failed to perform their duties.³⁴ In areas with weak state control over local administration, however, local administrators are more likely to be confronted with higher levels of threats against their lives, and less able to implement junta policies. In areas under complete control of non-state actors, the state’s local administration ceases to function, and is replaced by the representatives of non-state actors.

Figure 2. Analytical framework.

Our second set of explanatory framework focuses on factors that influence local administrators’ varied responses towards the coup in the strongest resistance areas which are still under military control or in areas contested by resistance forces. We show that local administrators’ varied responses are based on the nature and scale of their loyalty to the military regime, their relationship with local residents, and the evolving political situation. These varying responses have in turn shaped whether they are perceived as “agents of the state,” or “agents of the revolution” by local populations and the resistance movement. The behaviour of “agents of the state” and “agents of the revolution” also affect the nature and quality of local administration, as agents of the state are more likely to strictly enforce junta policies, provide survillience and crucial information about the resistance movement, and help strengthen junta control over local administration. Meanwhile, “agents of the revolution” can help undermine junta control over local administration through actions that protect and promote the resistance movement.

The local state administration in A, B, C remain relatively intact and able to perform basic or limited-basic functions due to the lack of strong resistance against the SAC. These areas for instance experience lower level of armed clashes and assassination of people accused of supporting SAC or anti-SAC groups. Examples of category A include the SAC-controlled areas in Naypyitaw, Ayeyarwady, some parts of Yangon, Mandalay, and in major towns and army bases in provincial areas across the country. Examples of category B include areas that are shared between SAC and ceasefire Non-State Armed Groups (NSAGs) such as the Pao National Organization and Kayan New Land Party in Southern Shan States, and Border Guard Forces in Shan, Kayin/Karen and Kachin states. Examples of category C include contested areas between SAC and the Arakan Army during the ceasefire period (November 2020–June 2022; November 2022–October 2023).

The degree of state control of local administrations in categories A, B, C is relatively higher, shown in terms of local administrators’ ability to perform some “basic” functions such as implementing and enforcing SAC policies and ensuring compliance on payment of utility fees, taxes, and renewal of vehicle licenses. WA/VTAs are able to visit households to check on visitors and to collect household lists. They also provide some limited services such as issuing recommendation letters for travel, passports, and job applications. However, they have not been able to deal with the rise in crimes, thefts, robberies, and corruption associated with the breakdown of law and order. There is a lower corresponding rate of resignation and assassination of WA/VTAs. Local administrations with “limited basic functions” (particularly in areas with overlapping hostile groups) have a lower level of state’s enforcement and service provision, and higher level of resignation and assassination of WA/VTAs.

There is no state local administration under the areas controlled by NSAGs in D and H. In categories D and H, local affairs are either managed by resistance groups or a collective body of community leaders, or by traditional leaders.³⁵

Empirically, there is very little resistance against the SAC in mixed areas controlled by SAC and SAC-friendly NSAGs (F) because the latter which controls the ground is able to suppress anti-SAC resistance. In overlapping areas controlled by SAC and NSAGs friendly to the SAC (area B), local affairs are run by either one representing both SAC and NSAGs, or two separate local authorities representing the SAC and NSAGs respectively. In Rakhine state, for instance, the Arakan Army, which signed a ceasefire with the SAC in late 2020 (and again in late 2022 until the end of October 2023), used the SAC’s local village leaders and administration to channel state resources (such as agricultural loans) to communities under contested areas. It either collaborated with or subsumed local SAC bureaucrats in some areas, or ran its own village committee parallel to the SAC’s.³⁶

Some wards in SAC-controlled areas in Yangon³⁷ and Mandalay regions (E), and in contested areas such as Sagaing and Magway (G) face strong anti-SAC resistance on the ground, leading to higher levels of resignation and assassinations of WA/VTAs, and the near collapse (“barely functioning”) or collapse of local state administration. The Ayeyarwady news outlet reported in September 2021 that over 100 local administrators in Sagaing, Magway and Yangon regions resigned from their posts in response to threats from the resistance movement.³⁸ The highest number of local administrators were killed in three Bamar-dominated regions: Mandalay (139), Yangon (110), Sagaing (78), Tanintharyi (53) where NLD support and anti-junta resistance are the strongest (Figure 3).³⁹

Figure 3. Variation in state control over local administration.

In some places in strong resistance areas, no individual was willing to assume the position given threats and pressure from both sides.⁴⁰ In an urban town in Kayah state, one community leader was approached by his co-residents to perform the tasks designated for WA/VTAs.⁴¹ In one village in Tanintharyi, the SAC forced three leaders from each neighbourhood (a total of 35 residents in the entire village) to serve as a collective administrative body.⁴² In other situations, the dominant authority is whichever stronger armed group is able to have their local representative run local affairs, or a WA/VTA ends up representing both SAC and the resistance group, or two WA/VTAs end up representing each organization separately. For instance, there were reportedly two local representatives for Taang National Liberation Army and SAC respectively in Northern Shan state (Figure 4).⁴³

Figure 4. Number of assassinated local authorities by state/regions (From 1 February 2021 to 22 September 2023). Source: Armed Conflict Location and Event Data (ACLED). Myanmar, 2021–2023. https://acleddata.com.

It should be noted that while state local administrations barely function in both places, the nature and scale of violence differs in these two types of control areas. Anti-SAC resistance movements operating in the SAC-controlled areas (E) had to rely on guerrilla attacks and targeted bombing of buildings and assassinations of alleged SAC supporters. In the contested (G) area, however, the SAC also has to fight hostile NSAGs who contest for control, which are manifested in terms of direct armed clashes, airstrikes and shelling and burning of villages by SAC, in addition to guerrilla attacks and targeted assassinations by the resistance movement. Generally speaking, NSAGs are able to collect taxes, and provide limited services such as security, education, and dispute settlements in contested areas (G) than in SAC-controlled areas (E).

Variations in local authorities’ responses in high resistance areas (E, G)

We focus mainly on three regions, Yangon, Mandalay, and Sagaing, which have experienced the highest number of assassinations of local authorities by the resistance movement, to demonstrate a variety of strategies adopted by different local authorities who confront similar sets of risk, and must balance their acts between competing armed forces to reduce threats against their lives.⁴⁴All of these places are Burman-dominated regions and new resistance areas which enjoy higher level of support for the NLD. Out of 44 townships that have been under martial law since the coup, 14 of them are in Sagaing, 6 are in Yangon, and 5 in Mandalay.⁴⁵ Yangon and Mandalay are hosts to two major cities with a growing mixed of migrant populations from across different ethnic and regional backgrounds. Sagaing has been predominantly Burman, but it has several enclaves occupied by ethnic minority groups such as Chin, Naga, and Shanni. Data below show that assassination took place in all townships that were under martial law in Yangon (Hlaingtharya #13, Dagon Myothit (Seikkan) #7, Dagon Myothit (South) #6. Shwepyithar #4, Dagon Myothit (North) #4, North Okkalapa #2) and several townships that were under martial law in Sagaing regions (Kawlin #5, Tantse #4). None of the assassination occurred in martial law townships in Mandalay (Figure 5).

Figure 5. Number of assassinated local authorities by townships in Yangon, Mandalay, and Sagaing (From 1 February 2021 to 22 September 2023). Source: Armed Conflict Location and Event Data (ACLED). Myanmar, 2021–2023. https://acleddata.com.

The situations vary across different townships in each region, depending on the proximity to the town, topography, and political affiliation and socio-economic backgrounds of residents. As of 22 September 2023, most of the townships in Yangon and Mandalay were under the SAC’s control. An estimate of eighty percent of townships in Sagaing was either contested or administered by the new resistance groups, except for a couple of major towns that are under the SAC control. Even in some SAC-controlled areas such as in Yinmabin, residents reportedly need to get letters of recommendation for travel, property sale, or job and passport applications from township authorities since village administrators stopped providing services despite allegedly continuing to receive state salary.

Many local administration offices in the post-coup period in these strong resistance areas were deprived of resources and become vulnerable to attack by the resistance movement. One resident in Yangon said:

I went to the ward office to register my son’s birth, but I don’t want to stay long because it is a dangerous place. This office used to be lively and crowded. The ward leader used to be assisted by clerks, but now he does everything on his own. He complained about being overwhelmed with many urgent needs, including provision of security for schools, to taking actions against theft cases.⁴⁶

WA/VTAs who were caught in between and are confronted with the same set of risks and dilemmas therefore must make difficult decisions on whether they should keep or accept the position, whom they should side with, how to balance acts among all rival armed groups, and how to mitigate risks against their lives.

Because the situations remain fluid, and the nature and intensity of violence evolve and vary across time and space, we find that variations in local authorities’ responses to the coup-related crisis are influenced not only by the nature of their relationships with the SAC authorities, with their constituents, and with the resistance movement, but also by their strategic calculation based on evolving local and national situations.

One factor that influences their decision and practice is the nature of their political affiliation and the level of their ideological commitment to military rule. Pro-SAC WA/VTAs are more likely to remain in office than their anti-SAC counterparts. For instance, immediately after the coup, many WA/VTAs appointed under the NLD government (2011–2016) resigned in protest. The SAC either replaced the vacancies or removed local authorities loyal to NLD with SAC appointees. Initially, ethnic minority states seem to be less affected by SAC reshuffling as well as targeted assassination of local authorities by the resistance movement than Bamar-dominated regions, most probably because of the presence of ethnic minority political parties and armed groups, some of which have developed antagonistic relationships with the NLD or have established their own local administrations. According to a survey conducted by Nyan Corridor immediately following the coup in February and March 2021, 73% of respondents in ethnic minorities states reported no change in local administrator, compared to 12% of respondents in Bamar-dominated regions.⁴⁷

Newly appointed unarmed administrators are usually warned by resistance groups to resign from their posts. Their lives are spared if they resign, but they became “soft” targets for assassination if they do not.⁴⁸ A local researcher from Mandalay observed: “One ward office got bombed but the local administrator continued to work from home. Soon a person who came to get a recommendation letter at his house killed him.” She also said:

The resistance movement will continue to plot assassination until it is successful. They would use any kind of weapons that they can get. Someone used bicycle spokes to stab a ward head, who is now getting treatment and is in hiding.⁴⁹

Our preliminary study based on 26 surveys shows that SAC was more likely to appoint and retain leaders who were either pro-SAC (who are generally ex-military officials, members of USDP, or leaders and members of SAC militias such as pyusawhtee) or who displayed no political affiliation. For instance, most new appointees (six out of nine) in 26 places are USDP members, ex-military leaders or members of SAC militias.⁵⁰ One is an NLD member, and two belong to parties that are neither USDP nor NLD. Of local authorities who have retained their position from the NLD period, one is pro-military (USDP), three are NLD, four are from ethnic parties, and five others have no known political affiliation. This shows that while many NLD local authorities resigned in protest or are removed by the SAC, some of them retained their positions. All NLD who retained their position immediately following the coup in fact are from ethnic minorities areas.

Our survey also shows that WA/VTAs with pro-SAC leaderships presiding over predominantly NLD constituents tend to have the most hostile relationships, while leaders with no political affiliation or those elected by the community tend to have friendly relationships with their residents. Researchers describe the relationships between pro-SAC leaders and predominantly pro-NLD residents as “hostile” (Mandalay1) or “neither friendly nor hostile” (Yangon10, Sagaing2).⁵¹ A respondent from Sagaing15 ranked the relationship between local leaders and residents as “good” in a place where village affairs were collectively managed.

These findings are further supported by written responses from separate interviews conducted in Sagaing and Yangon. One respondent from Sagaing said: “Social relationships suffered when NLD supporters became resistance members and USDP supporters took charge of local administration.”⁵² Another respondent from Sagaing said:

The wife of a village head in a nearby village gathered relatives, recruited and offered cash to poor people to attend (military) training by pyusawhtee (a pro-SAC para militia). She accused people against whom she had a personal grudge as being resistance members so they were arrested and tortured by the army. That’s why she was killed by local defense forces … . This kind of situation doesn’t occur in villages whose leaders were not appointed by SAC. Villagers don't even argue and fight among themselves. If there were tensions, elders would guide and mediate problems.⁵³

The level of willingness to serve the SAC, however, varies among individual WA/VTAs who are generally described as pro-SAC. Some reluctantly retained their positions due to their inability to resist SAC, while others willingly serve the SAC out of deep ideological commitment or to utilize the office as a venue to enhance their power and financial interest. Political Insider noted the opportunity created by growing lawlessness for petty corruption and extortion allowed local administrators to overcharge license fees from local businesses and service fees from local residents, and to extort money from local residents for minor violations of laws.⁵⁴

Variations in behaviour and practices of WA/VTAs towards the SAC, local constituents, and resistance movement also influence the chances of their survival (either remaining alive or retaining their position) and demise (dismissal by the SAC for “failing to perform their duties,” or assassination by the resistance movement). These outcomes also have an impact on their future decisions to resign or modify their behaviour and practices.

Our preliminary study finds that WA/VTAs who publicly and openly sided with the SAC authorities are generally regard as “agents of the state” by residents and tend to be the target of assassination by the resistance movement. Corresponding behaviour of “agents of the state” include providing crucial information about resistance movement, escorting security forces to inspect households allegedly supporting the resistance movement, helping make arrests of revolutionary members, fully implementing SAC rules, abusing power and authority, and extorting money from residents. A respondent living in Yangon for instance said:

The new head appointed by the SAC was a taxi driver who has good public relations skills and got along with his residents, but soldiers were seen visiting his office and three local young people were arrested. So people increasingly became unhappy with him and a bomb exploded next to his office.⁵⁵

Some leaders who were killed reportedly went to great length to abuse their power:

A local head was killed for handing in young people into SAC hands. A while ago, he and the army ordered two young people who sang resistance songs to sit on their knees for the entire night and slapped their faces.⁵⁶

Respondents also commented that corrupt local heads were more likely to be killed; “they collect electricity fees” (which is not the responsibility of WA/VTAs).⁵⁷ Political Insider also wrote that underground groups have often sought to justify assassinations by claiming the victims had been extorting residents.⁵⁸

On the other hand, WA/VTAs who avoid full implementation of SAC policies, engage in activities that are perceived as protecting local residents and members of the resistance movement, “get along with all sides of armed groups,”⁵⁹or “keep a low profile,”⁶⁰ are identified as “agents of the revolution” and tend to avoid the risk of assassination. They are also credited for helping prevent or mitigate potential violence. Protecting local residents includes warning resistance groups of forthcoming searches and attacks or bailing out villagers detained by security forces, or not informing security forces.⁶¹ All these valued qualities are summed up for a VTW in xxx township in Yangon who served for 20 years and who

knows the residents very well, who keeps very low key, and does not cause trouble to people. He does not do exactly what he was ordered by SAC to do. He initially planned to resign, but he still has to serve because they could not find his replacement.⁶²

Another resident from a different ward in Yangon said: “Our new ward leader is appointed by the army, but we understood him. We need his witness signature for sale contracts or transfer of wills, and his recommendation letters to apply for jobs or passports. He also has good relationships with community members because his wife runs a local diner which serves community members.”⁶³

Some leaders who got along with their constituents were regarded as highly respected figures, and elected by the community (Sagaing10) or reelected over different governments (Yangon14). A researcher in Mandalay said of her newly appointed village head:

Community members don’t cause him trouble because he used to be active in community social and religious affairs. He is a USDP member but he did not show any overt support for it. His wife was against his appointment, but he was also afraid he would be in trouble if he did not take the position.⁶⁴

One village head in Sagaing also said:

I did not want to be a village head after others resigned. But under too much pressure and repeated requests from the community, I accepted it. I had a meeting among ethnic minority leaders immediately after I was elected; I told them they could support whoever they want as long as they do not harm religious harmony, education or community peace.⁶⁵

Several interviewees from predominantly pro-NLD neighbourhoods administered by SAC appointees emphasized WA/VTAs’ ability to maintain the balance between the two opposing forces as a key factor in determining peace and conflict.⁶⁶ Two respondents estimated that 50–60% of peace and conflict is influenced by the role of local administration and said that village leaders’ wisdom and clever approach (panat lanemar mhu) are important in preserving peace within the community.⁶⁷ “He/she should be able to deal with both sides. Residents’ sufferings can be ameliorated if he/she thinks for the neighborhood welfare.”⁶⁸

Inability to balance acts sometimes comes at a cost. One resident in Yangon said of his ward administrator: “He became a monk when the SAC authorities came to our ward to appoint a new administrator and refused to accompany them when they came to search houses in the neighborhood.”⁶⁹ In 2022, this ward leader reportedly had a heated argument with military officials when he was forced to recruit soldiers from his ward.

Generally speaking, we find that respondents’ descriptions of “agents of the state” and “agents of the revolution” are consistent with narratives used by the resistance movement in social media regarding why certain individuals were killed. Of course, not all members of local communities share and accept the views and practices of local resistance forces (known as People's Defence Force PDF). Local communities are divided by socio-economic status, political affiliation, generational, ethnic, and religious differences, and have competing interests, while local elites include formal leaders such as local officials and traditional leaders as well as informal leaders such as religious leaders, businesspeople, and intellectuals. Some a WA/VTAs in our surveys have a sizable presence of residents from multiple ethnic and political affiliations, the effects of which on relationships with village heads need to be further assessed. In Sagaing2, where USDP (the military’s affiliated party) members constitute 55%, and NLD constitute 40%, for instance, village unity was maintained when strong supporters of both sides fled, leaving people without extreme political loyalty to contain extreme violence. In addition, we also found that respondents sometimes make simple assumptions about local authorities: “If they don’t engage in the civil disobedience movement, they are dalan (informers) and SAC stooges.”⁷⁰ There is no way for us to confirm that individuals were killed for the alleged reasons. Some local leaders displayed all the characteristics as a candidate for assassination, but they remained unharmed due to the position of community members who do not want to cause disturbances or in the absence of strong resistance. However, given the NLD’s overwhelming election victory in 2020, and nationwide resistance to the coup, many civilians and resistance movement share similar views regarding the SAC.

Finally, we find that many local authorities have adapted themselves to evolving political situations. A local researcher in Yangon who ranked the relationship between a pro-SAC leader ruling over predominantly NLD residents as “neither friendly nor hostile” remarked on a local head’s changed behaviour after his predecessor and several local authorities were assassinated. They were accused by PDFs as being SAC informants and held responsible for the shooting and arrest of young protesters against the coup:

The current SAC-appointed ward administrator does not see people, and barricades his house with two layers of metal fences. He closes his office by noon and lets 100 household leaders and clerks take care of most of the paperwork such as visitor registration.⁷¹

Another resident from Sagaing says the brother of a pro-SAC village tract leader, who was killed because “he was arrogant and threatened villagers,” was appointed by SAC to replace the position, but “he avoided the public and stayed quiet.”⁷² Some have taken precautionary measures by working from their homes or from safer locations such as Buddhist prayer halls and other religious buildings. Others bar residents from entering their offices and instead order them to submit and receive documents via openings in the sides of the building.⁷³

However, we also found a different kind of responses by WA/VTAs who survived attempted assassinations or witnessed assassination of their counterparts, depending on which armed authorities have more power in a specific location. A resident of Mandalay says: “In our neighborhood, the ward office was bombed, but the ward administrator was not killed. From then on, SAC stationed soldiers, guarded the place, and he continued to carry out his works.”⁷⁴ In addition, being removed by the SAC for failing to perform one’s duty does not necessary protect the life of one former administrator from the resistance movement. A resident from Sagaing recounted that a village head in another village was warned and later killed because he worked with the SAC to collect electricity fees. The respondent’s village leader, who also happened to be the brother of the assassinated administrator, is also a USDP member among overwhelmingly pro-NLD residents:

his life was spared, because he is a relative of all village community members. His children are pro-NLD and they discouraged him from serving as a village leader. He also keeps a low profile. Plus, our village is a small and close-knit community. He used to be arrogant and actively worked for SAC, but after his brother was killed, he avoided the public and stayed quiet.

As a consequence, he was later fired by SAC for “failing to perform his duties.” The same respondent, however, say in August 2023, this former village administrator, along with another former administrator from a different village was arrested by unknown armed groups (presumably PDFs) and reportedly were killed. “We don’t know who arrested him, but most probably because they still collected utilities fees on SAC’s behalf.”

Implications and conclusion

Our research attempts to close the gap between the literature on civil war and on authoritarian governance by demonstrating how the same kind of actor (local administrators who are part of the state administrative machinery) who face similar sets of dilemmas and risks associated with civil war can adopt multiple approaches to balance acts between opposing armed groups. In the first set of explanatory framework, we show that the type of local administration affects administrator behaviour to the extent that local representatives in areas of strong state control over local administration are more likely to have a higher level of confidence in implementing junta policies, and the junta is able to replace local administrators who resign from their posts or fail to perform their duties. While “agents of the state” and “agents of the revolution” exist in both junta-controlled areas and contested areas, we zoom in on local administrators’ varied responses towards the coup in the strongest resistance areas in our second explanatory framework set. We find that (1) local administrator's partisanship, (2) local administrator-resident relationship, and (3) evolving political situations help explain variations in responses by different local administrators towards threats against their lives.

More research is needed in terms of conducting large N samples to affirm the findings of our preliminary study, and offer nuances in terms of defining the criteria for the quality of local administration, and “hostile” or “friendly” relationships, and the nature and intensity of intra-village/ward splits.

Our tentative findings nevertheless offer some general conclusions about the importance of local administrators in deepening or undermining the resilience of authoritarianism. In highly polarized and violent contexts, local authorities’ ability to mediate conflict, minimize violence perpetrated by competing armed groups, and to protect and advocate for their local community are highly valued assets. They can help enhance military dictatorships by refusing to step down, or by strictly implementing military rule (especially in SAC-controlled areas) or by cracking down on the resistance movement. Some of these local leaders who show no strong political affiliation, or who are highly respected individuals, or who attempt to protect local populations and resistance movements, have been able to ameliorate the worst forms of authoritarian practices. Although there are other broader factors that affect the conditions of peace and violence in the local context in post-coup Myanmar, the often-neglected behaviour and practices of local leaders in conflict-affected and authoritarian contexts can help make a difference between life and death.

Acknowledgments

The authors would like to thank Nhu Truong, Elvin Ong, and two anonymous reviewers for their valuable insights and constructive feedbacks. We are grateful for colleagues who provided helpful comments at various conferences and workshop where we presented our papers. These include Association of Asian Studies, Burma Studies Conferences, Southeast Asian Brown Bag seminars at Yale University, University of Michigan, and Brown university. All errors in this paper are ours.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by International Peace Research Association Foundation and LIFT MYANMAR.

Notes on contributors

Ardeth Maung Thawnghmung

Ardeth Maung Thawnghmung is professor of political science at university of Massachusetts Lowell.

Naw Moo Moo Paw

Naw Moo Moo Paw is a PhD student of Global Studies at the university of Massachusetts Lowell.

Notes

1 Kempel and Tun, Myanmar Ward and Village Tract Administrator Elections 2016. Under an earlier military dictatorship (1988–2010), approximately 16,800 ward and village tract administrators were appointed by the General Administration Department (GAD) of the Ministry of Home Affairs. The Ward and Village Tract Administration Law of 2012, however, introduced indirect elections for Ward (urban), and Village Tract (rural) administrators under which a “secret ballot” is collected from 10 household heads. Most rural residents continued to rely on 100 household leaders, sometimes refer to as “village administrators,” as informal authorities, who assist village tract administrators. In this article, we focus on village (or 100 household leaders), village tract, and ward administrators who serve as both formal and informal authorities and conveniently refer to all of them as WA/VTAs. In 2017, the GAD was placed under civilian control, but it was again placed under military control following the coup.

2 Ibid.

3 Kalyvas, The Logic of Violence in Civil War.

4 Humphreys, “Natural Resources, Conflict, and Conflict Resolution,” 49; Fearon and Laitin, “Ethnicity, Insurgency, and Civil War,” 97.

5 Bertrand, Pelletier, and Thawnghmung, Winning by Process.

6 Rubin, “Rebel Territorial Control and Civilian Collective Action,” 64.

7 Collier and Hoeffler, “On the Duration of Civil War,” 41.

8 Mampilly, Rebel Rulers.

9 Arjona, Rebelocracy. Also see Rubin, “Rebel Territorial Control and Civilian Collective Action.”

10 Kaplan, Resisting War.

11 Van Baalen, “Local Elites, Civil Resistance, and the Responsiveness,” 58.

12 Kaplan, Resisting War.

13 Askew, “Landscapes of Fear, Horizons of Trust,” 40.

14 Idler, Garrido, and Mouly, “Peace Territories in Colombia,” 10.

15 Rubin, “Rebel Territorial Control and Civilian Collective Action”; Sarbahi, “Insurgent-Population Ties,” 47; Staniland, Networks of Rebellion; Wickham-Crowley, Guerrillas and Revolution in Latin American; Kaplan, Resisting War.

16 Buhaug, Cederman, and Gleditsch, “Square Pegs in Round Holes,” 58.

17 Collier and Hoeffler, “On the Duration of Civil War,” 41; Dube and Vargas, “Commodity Price Shocks and Civil Conflict,” 80; Humphreys and Weinstein, “Who Fights? The Determinants of Participation,” 52.

18 Larson and Lewis, “Rumors, Kinship Networks, and Rebel Group,” 72.

19 Owen, “Participatory Authoritarianism.” Zhao, “The Cadre System in China’s Ethnic Minority Regions.” Also see Lee and Zhang, “The Power of Instability.”

20 McClean, The Government of Mistrust; Chen and Xu, “From Resistance to Advocacy.”

21 Khuu, Jones, and Ekins, “Governance Analysis of Nha Trang Bay.”

22 Chen and Liu, “Village Leaders, China.”

23 Fox, The Politics of Food in Mexico. Also see O'Brien and Li, Rightful Resistance in Rural China. Thawnghmung, Behind the Teak Curtain.

24 Approved by the UMass Lowell’s IRB 21-101-THA-EXM in July 2021.

25 Approved by UMass Lowell’s IRB 22-071-THA-EXM in May 2022.

26 These local researcher are based in Yangon and Mandalay respectively, and have done extensive research on local governance.

27 Secondary sources include aggregate data compiled by the ACLED, social and conventional media, and NGOs reports. There is contradictory data on armed clashes, casualties and displacement from different organizations using different measurements on the same categories of observation. It has become increasingly difficult to verify data in an environment of armed conflict and repression. We use ACLED’s data due to its relatively consistent measurements over a longer period of time, and to contextualise an evolving situation, rather than presenting their conclusions as uncontested facts. In 1989, the Myanmar junta replaced existing English names for the country and its divisions, townships, cities, streets, and ethnic groups with what is considered to be the more authentic Myanmar names. Thus Burma became Myanmar, Rangoon became Yangon, and ethnic groups such as Karen were renamed Kayin. We will use Burma and Myanmar interchangeably in this article, but will use post-1989 names for states and regions to avoid confusion.

28 Thawnghmung and Htoo, “The Fractured Center.”

29 Thawnghmung and Noah, “Myanmar’s Military Coup.”

30 Myanmar’s administrative regions are divided into seven ethnic minority states, named for national minorities that constitute a majority in their respective geographic areas, and seven regions, which are predominantly Bamar. There is no reliable and current data on ethnic composition in Myanmar. Demographic changes have also transformed ethnic composition in areas that are traditionally regarded as exclusively minority areas or exclusively Bamar-majority areas. According to the 2014 Census, almost 9.4 million of the country’s 51.5 million people were internal migrants (Republic of the Union of Myanmar 2016). There were large population movements from poverty-stricken regions, where the Bamar are the dominant ethnic group, to less populous, more fertile, and resource-rich regions occupied by minority ethnic populations. Only Chin and Rakhine states, the two poorest ethnic minority states in Myanmar, suffered population lost among areas occupied by minority ethnic groups. Many ethnic minorities people have also moved to Yangon region and Mandalay region for better economic and educational opportunities. The 2014 census however did not provide any data on ethnic backgrounds due to the controversy over how to define ethnic groups and who belong to “indigenous” categories. Callahan, “Distorted, Dangerous Data,” 32.

31 ACLED, “Myanmar, 2021–2022.”

32 SAC-Myanmar, Briefing Paper: Effective Control in Myanmar.

33 On 27 October 2023, a coordinated attack by the Brotherhood Alliance against SAC outposts in Myanmar-China border areas sparked offensives against the SAC and expanded areas of violence. Within two months, the SAC lost control over several towns and hundreds of military bases across the country. Areas previously peaceful such as Lashio and many towns in Rakhine state have now been transformed into conflict areas (Hein, “Charting the Shifting Power Balance”; ACLED, “Myanmar, 2021–2022.”; Clapp and Tower, “Myanmar’s Junta Is Losing Control”). The findings of this research are based on the situation up to October 2023.

34 This is based on analysis of a comparison between high resistance and low residence in our 26 surveys and 50 interviews (Table 1). Twelve areas are described as strong resistance areas, thirteen are described as low resistance areas, and one was under the control of the resistance movement in 26 surveys. 28 locations are described as strong resistance areas and 22 are described as low resistance areas in 50 interviews.

35 Independent Research Network, “Understanding Hybrid governance in post-coup Myanmar (Part 1 and Part 2).”

36 Ibid., virtual Interview with a lead researcher on local governance, July 5, 2022.

37 These include some wards in townships that are under martial law in Yangon Region.

38 Irrawaddy, “Hundreds of Myanmar Junta Informants Killed.”

39 Noted that these figures also include township officials, who are not part of the article’s focus. Yangon region has the largest populations (over 8 million), but is geographically the smallest of all regions/states. Mandalay is one of the five smallest regions (ranked 10th out of 14 states/regions) but has the fourth highest population (over 6 million).

40 Reported by researchers from two townships in Sagaing.

41 Virtual Interview with head researcher on local governance, March 26, 2023.

42 Ibid.

43 Ibid.

44 Some of those assassinated local authorities in the ACLED data also include township authorities and village administrators and 100 household leaders.

45 https://www.aljazeera.com/news/2023/2/4/myanmar-military-expands-martial-law-in-strongholds-of-resistance; https://www.ispmyanmar.com/3551-2/; https://www.rfa.org/english/news/myanmar/martial-law-02032023190606.html.

46 Author’s phone conversation with a relative, March 10, 2022.

47 The Nyan Corridor, “Public Perception Study in the Post-Coup Myanmar,” 9.

48 Political Insider, “Motivating the Minions”; Irrawaddy, “200 Myanmar Junta Appointed Administrators Killed”; Mizzima, “Bullying’ Junta Village Administrators.”

49 Author’s phone interview, June 6, 2021.

50 Mizzima, “The Rise of the Pyu Saw Htee.”

51 Our survey is based on respondents’ perception and ranking of 3 levels of admin-resident relationship which are subjective, thus provide various, sometimes inconsistent, rationales for these numerical rankings.

52 Sagaing7, 2022.

53 Sagaing15, 2022.

54 Political Insider, “Motivating the Minions.”

55 Author’s Phone (Facebook Messenger) interview, August 4, 2022.

56 Yangon10, 2022.

57 Sagaing3, 2022.

58 Political Insider, “Motivating the Minions.”

59 Sagaing1, 2022.

60 Sagaing15. Yangon 11.

61 Yangon11, Sagaing 1, Yangon 19.

62 Yangon5.

63 First author’s phone conversation with a friend, September 6, 2021.

64 Author’s phone interview, June 6, 2021.

65 Sagaing10.

66 Sagaing1, Sagaing 5.

67 Sagaing5, Sagaing 6.

68 Sagaing5.

69 Yangon6.

70 South Shan2.

71 From 26 Surveys.

72 Sagaing2.

73 Political Insider, “Motivating the Minions.”

74 First author’s phone interview, June 6, 2021.