When authoritarian innovations fail: Labour governance in Myanmar's reform era and beyond

Abstract

The concept of authoritarian innovations provides a means of interpreting labour governance practices that claim to uphold democratic values and international labour standards while restricting labour participation and genuine voice in the workplace. Two underexplored dimensions of this literature are (a) how authoritarian practices are utilised by power elites in ‘hybrid’ state formations and (b) under what conditions they may be discarded in favour of a return to direct forms of coercion and violence. In this article, we examine these questions with reference to Myanmar, where there was a decade of experimentation with democracy before a reversion to military-dominated authoritarian rule in February 2021. In the first half of that decade, space opened under a military-led government for decidedly more democratic labour relations, although labour governance reforms were designed and operationalised in ways that constrained the strength and reach of trade unions. In the second half, a democratically elected civilian-led government presided over the contraction of the space available to organised labour. By tracking the alignment of meso-level authoritarian practices against the changing macro-level character of Myanmar's political landscape, this article contributes to theoretical debates within the emerging literature on authoritarian innovations in labour governance.

Introduction

The concept of authoritarian innovations provides a means of interpreting labour governance practices that claim to uphold democratic values and international labour standards but that in effect restrict labour participation and genuine voice in the workplace. Because they do not rely on direct forms of state coercion and violence, these innovations can occur within a range of regime types, from established democracies to one-party states (Curato and Fossati, 2020). Its attention to meso-level institutions and the operationalisation of governance practices (Pepinsky, 2020) make the lens of authoritarian innovations an ideal one through which to consider how the reform of employment relations institutions and practices designed and operationalised to constrain democratic participation at work can be falsely represented as enabling democracy.

Two underexplored dimensions of the literature on authoritarian innovations are (a) how these practices are utilised by power elites in ‘hybrid’ state formations, characterised by both authoritarian and democratic characteristics, and (b) under what conditions they may be discarded in favour of a return to direct forms of coercion and violence. In this article, we examine these questions with reference to Myanmar, a country historically defined by the centrality of military elite (Tatmadaw) in its political economy. Its decisive role in macro-level policymaking included a strategic decision in 2011 to experiment with hybrid governance and institutional reforms to enable greater domestic and international legitimacy. Myanmar's subsequent transition to a civilian-led government in 2016 – committed in its political platform to strengthening democracy, citizen rights and responsible economic development – was expected to build on the gains made in preceding years of reform in enabling worker representation and voice. But this was not the case. The National League for Democracy (NLD) government did not support Myanmar's nascent union movement nor address evident flaws in labour institutions that undermined unions’ credibility and effectiveness and collective representation. This contraction in the democratic space available to unions continued until the military again seized power in the coup of early 2021, terminating the experiment with a hybrid political regime.

As the Myanmar case demonstrates, there is strong analytical benefit in distinguishing between authoritarianism as a system of government and authoritarian innovations in a particular governance field – in this case, labour governance. We illustrate this benefit by first examining the expansion of democratic participation, labour rights, and collective representation via a series of reforms initiated by the military-dominated Union Solidarity and Development Party (USDP) government in an attempt to rebuild Myanmar's international legitimacy and reconnect with global investors and value chains. As our analysis underscores, these measures undoubtedly increased the capacity of workers and their representatives to participate in the workplace and in political life but were at the same time heavily tempered by the military's ongoing desire to maintain political control. We then turn our attention to the period of NLD rule, when a civilian-led government committed to progressing democratic reforms and institution-building at the macro level introduced policies and engaged in practices at the meso level of labour governance that in fact again reduced the space available to unions. This also included its failure to act by not pursuing required labour reforms or to reign in a labour bureaucracy that remained connected with the military and that was generally hostile to the exercise of worker voice and democratic participation.

Our analysis shows that labour governance reforms in Myanmar created space for workers’ democratic participation but were nevertheless designed and operationalised to constrain the strength and reach of unions. It also reveals that workers and unions were not passive victims during these periods of intense political and institutional change. Rather, they pushed back against authoritarian innovations during the ten-year window in which Myanmar inched towards democracy and then against the military's reversion to coercive authoritarian rule. In short, the Myanmar case illustrates the need for observing authoritarian practices that limit or close down voice and participation over time and how these practices align or misalign with the broader macro-level character of political regimes.

Authoritarian innovations and labour governance in hybrid regimes

Authoritarian innovations have been defined as ‘novel governance practices designed to shrink spaces for meaningful public participation in politics’ (Curato and Fossati, 2020: 1010). This concept, which originates in the political science literature, is distinctive because it moves away from a macro-level analysis of political regime types and electoral processes to understand how democratic participation, accountability and voice are stifled or subverted by often subtle changes in meso-level governance practices. Changes at the meso level can be achieved through a multitude of sometime subtle adjustments in regulation, bureaucracies, institutions or the strategic use of communication and political discourse to increase control. They can also manifest in a failure to act (Ford et al., 2021). More often than not, they consist of a ‘pattern of actions, embedded in an organized context, sabotaging accountability to people over whom a configuration of actors exerts a degree of control, or their representatives, by disabling their voice and disabling their access to information’ (Glasius, 2023: 22). Importantly, then, a series of authoritarian practices may be innovative in their combination rather than in their singularity.

The analytical benefit of this meso-level approach lies in its focus on the impact of specific changes in a single governance field across a variety of regime types. In other words, the concept of authoritarian innovations can be applied with equal utility in democracies where competitive electoral competition prevails, authoritarian single-party states or in hybrid political constellations that combine elements of both (Pepinsky, 2020). This is made possible by the fact that meso-level changes may coincide with – or contradict – a macro-level trend. On the one hand, a government may create space for (more) competitive electoral processes while simultaneously introducing authoritarian practices in a particular governance domain. On the other hand, authoritarian innovations can be introduced in a particular governance domain as part of a macro-level contraction of democratic space cloaked in the rhetoric of democracy.

Southeast Asia has been described as something of a ‘laboratory’ for meso-level authoritarian innovations (Curato and Fossati, 2020: 1013). It has also been noted as a region where labour governance and industrial relations institutions are influenced by the legacies of authoritarianism (Caraway et al., 2015). It is also a region where the contemporaneous influence of international organisations and global production network integration have conditioned state-labour relations (Ford and Gillan, 2016; Hess, 2021). While the region's economic dynamism is acknowledged by industrial relations scholars, little attention has been paid to the impact of political change and volatility on labour governance and patterns of state intervention (Ford and Gillan, 2016). In this context, there is a need for closer attention to how oscillating authoritarian and democratic tendencies in macro-level governance shape labour institutions and influence associated opportunities for voice and democratic participation. Although few studies examine these questions, some acknowledge that ‘authoritarian legacies’ continue to shape labour institutions and opportunities for democratic participation even in countries that have undergone a process of democratisation (Caraway et al., 2015). This strand in the literature recognises the likelihood of institutional continuities at the meso level even when broader macro-level political change has taken place. It also acknowledges the various motives and objectives – categorised by Hyman (2008) as supporting capital accumulation, maintaining legitimacy and pacification of labour conflict and demands – that inform state governance of workers and industrial relations (Brown, 2016; Crinis and Parasuraman, 2016; Ford, 2016; Ford and Sirait, 2016; Gillan and Thein, 2016; Hutchison, 2016; Ward and Mouyly, 2016).

International institutions, and the International Labour Organization (ILO), in particular, have played an influential role where countries have sought to expand their engagement in the global economy at a time when they are transitioning from deep authoritarianism to a more hybrid or democratic polity (Ford et al., 2018). In Southeast Asia, the ILO had a strong influence on the design of labour institutions in democratising Indonesia (Ford and Sirait, 2016), as well as in post-Khmer Rouge Cambodia (Ford et al., 2021) and newly independent Timor-Leste (Ford, 2016). Even unashamedly authoritarian Vietnam has been subject to international pressure in the form of conditions attached to trade agreements with the United States and the European Union (Marslev and Staritz, 2023; Tran et al., 2017). Yet, while international pressure has supported formal change, state authorities have a strong motivation to control the implementation of reforms. As a consequence, the power and impact of those reforms remain an open question (Anner, 2021).

If analysis of the implications of macro-level political change for labour governance in Indonesia is typically framed around the relative success of its democratisation process and in Vietnam around the efforts of an authoritarian state seeking to maintain domestic and international legitimacy while pacifying labour conflict, there is a greater degree of ambiguity and debate around the classification of regime types in other countries in the region. The limitations of binary classifications between liberal democracies and authoritarian regimes have long been debated in the political science literature, leading to the idea of ‘hybrid’ state regimes of various types as an alternative approach (Diamond, 2002). This distinction is especially relevant to developing Southeast Asia, which is home to a diversity of hybrid political regime types (Jayasuriya and Rodan, 2007; Lorch, 2021).

Hybrid states are defined by tensions and competing logics, including a desire for greater domestic and international legitimacy. Consequently, their emergence is often associated with political reforms that enable more democratic participation and a shift away from direct coercive forms of control. Greater distance from coercive authoritarianism can create space for democratic actors – including unions – that subsequently push for a deepening of democracy. Since ‘hybridity’ also points to a desire of political and economic elites to retain control, transitional governments often respond to these pressures by initiating institutional changes that maintain a veneer of democratic expansion but in fact constrain rather than enable genuine voice and participation. The labile state of these fragile democracies – indeed, the fundamental ambiguity of the prevailing political regime – makes hybrid regimes a particularly dynamic context in which to observe the regressive impact of specific shifts in labour governance even in the midst of an ostensible process of ‘democratisation’.

Cambodia, for example, is ‘neither fully democratic, nor strictly authoritarian but rather a mix or hybrid form of political system’ (Baaz and Lilja, 2014: 6). In the field of labour governance, the Cambodian state has both sought to enact reforms that afford it greater domestic and international legitimacy and that enable deeper integration with global production networks. But governing political elites continue to see workers and unions as a ‘problem to be managed through semi-authoritarian modes of governance aiming to maintain order and stability’ (Ward and Mouyly, 2016: 261). It is also a state where a dominant ruling party has refused to cede ground in politics and where macro-level authoritarian practices have become increasingly evident over time. For these reasons, meso-level authoritarian innovations in labour governance – for example, incremental changes in labour administration that serve to erode independent union representation – have been important tactical manoeuvres. Their utility lies in the fact that they allow for a ‘veneer of participation’ that allows the Cambodian government to maintain ‘the appearance of conforming to democratic principles and international expectations while purposefully narrowing the space available for discussion and dissent’ (Ford et al., 2021: 1267).

Myanmar presents us with an even more complex case. In the period of just a decade, it transitioned from coercive authoritarianism to a hybrid regime and then back to coercive authoritarianism. As such, it provides a compressed example of processes of regime change that generally occur over much longer periods of time. The task at hand, then, is to analyse successive macro-level regimes’ practices of labour governance over that decade. In doing so, we ask (a) how the emergence of hybrid governance enabled trade union rights and institutional participation but also authoritarian labour governance practices; (b) how a failure of democratic consolidation (under a democratically elected government) led to further authoritarian innovations; and (c) how unions were affected by the complete reversal of any semblance of democratic labour governance after the 2021 coup.

Context and method

Between 2011 and 2021, a hybrid regime type prevailed in Myanmar, characterised by a form of power-sharing between the military elite and democratically elected state officials. As Stokke and Aung (2020: 276–277) suggest:

Myanmar's mode of transition has institutionalized a hybrid (semi-authoritarian) form of rule where there are new and important democratic spaces, but where the substance of democracy is curtailed by constitutional regulations that guarantee the military positions of power and prevent substantive popular control of public affairs. ¹

Citizens and international observers understood that Myanmar was in the midst of substantial democratic reform. They also understood, however, that the design and implementation of these reforms were conditioned by the structural power of the military elite in the economy, state bureaucracy and the parliament.

The transition began with the establishment of a ‘reformist’ government led by President Thein Sein and the military-backed USDP, which projected a commitment to international economic and political reintegration. It sought to demonstrate this commitment by releasing political prisoners; negotiating with the NLD for its renewed participation in elections; introducing a raft of legal reforms to enable greater space for freedom of expression, rights, and more effective institutions; and engaging extensively with international organisations and governments to facilitate these institutional reforms (Crouch, 2017). After decades of isolation, underdevelopment and coercive authoritarian governance, the USDP's capacity to implement this reform agenda was patchy, as was its ability to address evident and continuing weaknesses in various policy domains, including education, health, and economic management. The depth and spread of its political and institutional reforms – which included a raft of new labour laws – was recognised by international organisations and other countries, leading to an influx of foreign investment and development assistance (Ford et al., 2017).

Just under five years later, the democratically elected NLD government assumed power. Its victory was met with widespread euphoria, but ultimately the NLD failed to consolidate Myanmar's democratic transition (Swe, 2021). This is perhaps not surprising given its centralised and personality-based leadership culture; its inexperience in government; and the fact that it had to contend with the military's ongoing presence in political and economic life (Egreteau, 2022). The NLD nonetheless remained the dominant party in electoral terms, if nothing else due to strong popular sentiment against the involvement of the military in politics (Swe, 2021). The results of the 2020 general elections reflected this continuing dominance in a near wipe-out of the USDP. While the military sought to blame this outcome on alleged electoral irregularities (Kipgen, 2021), its grip on both politics and the state bureaucracy was clearly weakening.

It was in this context that the military elite staged a coup in February 2021. A military-controlled administration was installed, a state of emergency declared, and elected parliamentarians arrested. The coup was met by a remarkable upsurge of popular resistance, explained by – among other factors – the extension and growth of democratic values and norms into civil society after ten years of quasi-democratic governance and reforms (Stokke, 2023). Notably, unions and workers played a leading role in staging strikes and protests in industrial hubs and urban areas (Ko Maung, 2021). However, all forms of resistance to the rollback of democratic voice and participation were met with repression and violence by the military. The Junta's coercive repression of dissent and widespread violations of fundamental labour standards and union rights in Myanmar after the 2021 coup prompted an ILO special commission of enquiry, established as a precondition for taking further measures to sanction the military government (ILO, 2023a).

While acknowledging these macro-level political developments, our attention in this study is on meso-level labour governance. We are interested in how democratic participation was both enabled by formal legal and institutional reforms – not least the capacity for independent unions to be registered and represent workers in disputes resolution, wage determination and tripartite bodies – but also restricted by the design and administration of these new labour institutions. We collected primary data during two fieldwork visits during the USDP period (2013, 2014) and two during the NLD period (2019, 2020), and through supplementary interviews conducted by internet-supported communication channels in 2021. ² Interviews were conducted in either Burmese or English, based on the preferences of informants, who were interviewed either alone or in small groups, and were typically between one to two hours in duration. We interviewed 85 people in the first two rounds of data collection and 98 in subsequent rounds.

Our interview subjects included elected union officials at enterprise level; officials at the federation and confederation level; and representatives of labour non-governmental organisations (NGOs), international brands, and employer and industry associations. We also interviewed Myanmar-based staff from relevant international organisations (the ILO, development agencies, and trade union Solidarity Support Organisations), as well as members of the National Arbitration Council and officials from the Ministry of Labour including three Directors General. In addition, we visited five factories in key industrial zones in the hinterlands of Mandalay and Yangon, and engaged as participant observers at union meetings, in an ILO workshop that brought together representatives of all the major union federations with employers and senior personnel from the Ministry of Labour; and at the founding congress of the Confederation of Trade Unions of Myanmar (CTUM). We also amassed an extensive collection of media articles and relevant grey literature including documents and reports produced by unions, NGOs, employers and the ILO's Liaison Office in Myanmar to enrich and validate our primary data.

Our collection of data via repeated rounds of fieldwork is consistent with a tradition of historically informed employment relations analysis (Forde, 2023; Patmore, 2018). Repeated interviews with our research participants, especially representatives of union organisations, allowed for a longitudinal and ‘slow’ and contextualised understanding of the research domain (Almond and Connolly, 2020). Our analytical approach in this study is consistent with ‘temporal bracketing’ which has been described as a ‘strategy for dealing with diachronic process data, that is, case study data that are composed of detailed event histories over time’ and which ‘involves decomposing time lines into distinct phases where there is continuity in activities within each phase and discontinuity at the frontiers’ (Langley, 2009: 919). Consistent with this approach, we assessed our data set with reference to key domains of labour governance: wage determination, disputes resolution, labour administration and freedom of association. Guided by the emphasis within the literature on the impact of meso-level governance practices on opportunities for democratic voice and participation, we then examined these data set categories to identify distinct phases of expansion and contraction of rights and democratic participation across two distinct periods of political administration.

A qualified expansion of democratic space

The transformation that took place between 2011 and 2015 was in many respects both fast and impressive. The reforms that made it possible for unions to register and operate at the workplace level and in tripartite labour institutions – clearly democratic measures in a country where unions had been altogether banned and workplace rights near non-existent –were a component of a broader program of political and institutional change initiated by the USDP government from 2011. Significantly, they were also an essential precondition for the easing of international trade sanctions, the increased internationalisation of the economy and greater integration with global value chains. This was especially the case for brand-based garment manufacturing, with its demands for social compliance reporting and at least minimal alignment with international labour standards (Ford et al., 2017). ³

The first, and arguably most fundamental, change to be implemented in labour governance was the passage of the 2011 Labour Organisation Law, which recognised the right of workers to form and register independent unions (and of employers to form employer associations). The passage of the Labour Organisation Law engendered confidence that activists could at least try to begin organising and representing workers, while Associated Rules released in 2012 opened a pathway for the registration of ‘basic’ (enterprise-level) unions and for collective agreement making, as well as setting out the procedures for establishing union federations and confederations (ILO, 2017).

At workplace level, legalisation of unions created a new sense of possibility about democratic voice and participation that had not existed previously. According to a worker activist who joined an enterprise union during this early period:

In the past, we didn’t even have a chance to talk, just like what we are doing now. We only had thoughts of how to get employment so that we can survive. We simply didn’t know what rights we had been missing. Now, things are changing with reform process. (Interview, January 2013)

The union subsequently made demands for increased wages, access to a medical clinic, overtime pay and for the employer to provide workers with social security cards for proper registration.

At the same time, the design of Myanmar's new industrial relations institutions incorporated substantial constraints on unions’ ability to organise workers, influence wage determination or engage in disputes resolution. ⁴ Workers by no means had access to full freedom of association. This was hardly surprising since the Ministry of Labour remained dominated by former military personnel, and local township labour officers were often ill equipped to deal with the demands of a reformed labour governance system. As another labour activist explained, ‘Initially, police and military intelligence kept watching me wherever I went’ (Interview, January 2013). While this activist and his colleagues had been able to register a union, corruption among township labour officers and hostility still defined many of their interactions with the state administration. In one case that we encountered, factory workers dismissed by their employer were subsequently attacked by thugs and blacklisted, despite the fact that the National Arbitration Council had determined that the employer's initial action was unjust and that the workers were entitled to reinstatement (Interview with dismissed workers, January 2013).

The subsequent growth in union numbers was nevertheless significant. By early 2017, 2341 basic unions and 136 township unions had been registered, and 20 regional federations, eight national federations and one confederation recognised (ILO, 2018). ⁵ The process of union formation was challenging both because of weak awareness and knowledge among workers on the proper registration and functioning of a union but also because of an inherent resistance from employers and the state bureaucracy to the idea of worker voice and institutional labour representation. As noted by the General Secretary of what was then the Federation of Trade Unions of Myanmar (FTUM) ⁶ :

It’s going to take time. We are in some kind of changing of the culture, because when there’s political transition, then there are the people, the bureaucrats, who don't want to change … And who don’t want to agree that there has been change. And the unions are now part of the, let’s say, the dialogue partners. So, this is a culture shock that we’re feeling on the ground. We’re seeing on the ground. (Interview, January 2014)

The new system required the registration of basic unions in 10% of enterprises in a given occupation or industry in the same geographic area (township) and then at regional and national levels before an industrial federation could be formally established. The degree of difficulty associated with registration was even higher for confederations at national level, which required the affiliation of 20% of all union federations. Notwithstanding the formation of farmer unions in these early years of reform (Gillan and Thein, 2016), the institutions associated with this system failed to extend legal protections and opportunities for voice and participation to workers in the informal sector (Campbell, 2019; De et al., 2020).

In short, the potential for union voice and representation at the level of the factory and the workplace expanded, but the likelihood of strong coordinated union voice at the industry level and in national institutions was constrained. Myanmar's emerging union movement was also characterised by a tendency towards ‘factionalism and inter-union conflict and competition’ (Gillan and Thein, 2016: 283). In particular, the capacity of the FTUM/CTUM to channel international trade union aid and support was contentious (Gillan and Thein, 2016: 281): some labour activists and unionists held that this officially registered union confederation was unsupportive of strikes and misguided in its attempts to use the official disputes resolution system (Maung, 2019). Yet, despite frustrations with the limitations of the system – and with each other – there nonetheless remained some sense of hope that it could be refined. As one national union leader suggested, ‘We see the space, so we’re trying to work within the space and then trying to push to another level’ (Interview, January 2014).

Complementing the Labour Organisation Law was a series of other legislative measures that opened space for worker voice. Key among them was the 2012 Settlement of Labour Disputes Law, which established procedural rules for taking industrial action and a system for labour disputes resolution, consisting of tripartite disputes resolution bodies and arbitration councils at the township, regional and national levels. Workplace coordination committees comprised of employer and worker representatives served as the first point of potential disputes resolution. Disputes progressed to tripartite township and regional disputes resolution bodies, with a National Arbitration Council serving as the highest level of appeal.

The practical challenges of operationalising this system were immense given the weak foundations of industrial relations and conciliation and arbitration expertise in Myanmar after five decades of military rule (Interviews with National Arbitration Council members, January 2013). These flaws were evident not only to unionists and other labour activists, but also to ILO officials (Interviews, 2013 and 2014). They included the fact that worker representatives on the workplace coordination committees were often nominated by employers rather than by workers. There was also a lack of clarity, consistency, and recorded detail in the determination of the higher-level bodies. And, most of all, these bodies had weak enforcement powers, and were largely ignored by non-compliant employers. There were also practical, procedural difficulties in engaging in legal strike action, resulting in series of protests and strike actions outside of the legal framework. As explained by an activist labour lawyer representing workers in an industrial zone in greater Yangon, in many cases channelled through the disputes resolution system, ‘Workers won the case but lost their jobs’ (Interview, January 2014). ⁷

In addition to these central legislative pillars, a raft of other new laws relevant to labour were passed, including the 2011 Right to Peaceful Assembly and Peaceful Procession Law; the 2012 Social Security Law; the 2013 Minimum Wage Law; the 2013 Employment and Skills Development Law; the 2015 Rights of Persons with Disabilities Law; the 2016 Shops and Establishments Law; and the 2016 Payment of Wages Law (ILO, 2023b). Of these, arguably the most impactful was the 2013 Minimum Wage Law, which introduced a minimum wage system premised on tripartite consultation. Unions were quick to seize on the opportunity to exercise voice through public campaigns and commissioned their own surveys of worker cost of living expenses to exert influence and pressure the government to advance the process and make a wage determination (Interviews, January and December 2014). In 2015, the process culminated in the first national minimum wage. In the face of vociferous opposition from employer and business associations, a national standard was set at 3600 kyat (AUD 4.50) per day. A subsequent round of union campaigns and cost of living surveys resulted in an upward revision of the minimum wage to 4800 kyat per day in 2018. While the wage determination process was clearly contested and unions remained concerned about its uneven implementation and enforcement, these outcomes were widely recognised as delivering significant gains for workers (Nguyen et al., 2021).

Following this initial expansion of rights and opportunities for participation, there was a gradual contraction of the available space to organised labour. The rapid process of labour reform in these years stretched the resources of the state bureaucracy. It also failed to erase deeply seated authoritarian practices and logics in the administration of these new institutions. Yet, despite these deep flaws, the passage of these laws sparked a never before-held belief among workers that they could take strike and protest actions to demand improved wages and conditions (Interviews with unionists and NGO activists, 2013, 2014). These public demonstrations and wildcat strike actions were often more effective in achieving positive outcomes for workers than the official system (Interview with union officials, January 2014). At the same time, persistent institutional failures – and the apparent unwillingness of the state to engage in substantive further rounds of reform – prompted unionists to be cautious. As a unionist who had been dismissed by a hostile employer explained, ‘I accept the fact that we now have more freedom to talk and to organise, but I think this democracy is a trap for us … We are earnestly waiting for more mature democracy’ (Interview, January 2013).

The contraction of space under a democratically elected government

There was already a sense of stalled momentum even before the USDP government was defeated at the polls in November 2015. In the assessment of the ILO's Chief Liaison Officer, the government retained an interest in maintaining basic compliance with international labour standards, but the rapid easing of international trade sanctions had decreased the ILO's ability to push the bureaucracy further (Interview, January 2014). For these reasons, the ILO was something of a bystander rather than the influential actor it had been in the early years of reform and, with its attention drawn to supporting tripartite national dialogue and donor-funded project work, not well positioned to support unions and civil society organisations to exercise the practical exercise of worker voice.

Despite the constraints inherent in the military's reservation of one-quarter of all seats in the national parliament and retention of control over selected ministries, the NLD's victory offered hope for deeper democratic participation in labour governance (Interview with member of the NLD Labour Committee, December 2014). In reality, however, the NLD presided over the erosion of union representation and democratic participation in the workplace. According to a global brand representative with direct experience of sourcing from Myanmar, the NLD's approach raised questions about which model of industrial development the country was embracing:

In terms of how you set these systems, are they looking towards the [European Union] or are they looking towards China? I would say that they are probably looking more towards China, to control things rather than opening up for the parties to agree. The trickiest thing is how you convince everyone that this is the best way forward. (Interview, January 2019)

There were a number of reasons for the contraction of the available space under the NLD. Chief among them was the fact that newly elected parliamentarians did not prioritise industrial relations and labour issues; in fact, they proved to be ignorant or even hostile to the notion of union representation and independent democratic participation in the workplace. According to the General Secretary of the Myanmar Industry-Craft Service Trade Unions Federation (MICS) – who had participated in lengthy multi-year tripartite deliberations on the industrial relations framework – tripartite recommendations for institutional and legal reforms had been subsequently disregarded and subverted in the national Parliament. In his words, ‘They just do what they want. They don't listen to unions’ (Interview, January 2020). These views were echoed by officials from other union federations, one of whom pointed to the ILO's weakening capacity to progress institutional reforms and less intense international pressure regarding labour rights as factors:

What I see is that under the previous government, the labour minister, the Director General … were under pressure for many years by the ILO and international community. But now the government and parliament members … do not really understand. That's why parliament members don’t consider the ILO standards. They said that is the ILO standard but we have to work according to Myanmar conduct. That's our problem with our parliament members. (Interview with CTUM affiliated union federation leader, January 2019)

The same unionist went on to note that the failure of the government and politicians to understand and engage with labour issues meant that senior labour bureaucrats were ‘doing whatever they want’. ⁸

Labour NGO activists and Solidarity Support Organisation staffers based in Myanmar shared these concerns (Interviews, 2019 and 2020). As a Solidarity Center staffer observed, ‘The NLD does not see labour as a priority, or even the economy as a priority, and looks at established labour unions with some sort of scepticism’ (Interview, January 2019). Meanwhile, as noted above, the labour bureaucracy in Myanmar remained largely under the influence or control of personnel affiliated with the military. These officials were understandably concerned with containing industrial conflict but also with labour's potential engagement in the political sphere. Workers and unions were thus caught between the control of an entrenched labour bureaucracy and the indifference, or even hostility, of democratically elected politicians.

This tension played out in the experience of workers and union representatives, who reported new procedural rules changing thresholds of the minimum number of workers needed to register basic unions, requiring unionists to verify their identity by means of photographic identification, and unions to submit annual financial reports prepared by a certified auditor (Interviews, 2019 and 2020). In addition, unions were given shorter time frames for re-registration. According to a CTUM official, this meant that even unions with strong membership and financial systems risked failing to re-register. The same official noted that labour officials were preventing new unions from registering by failing to provide the necessary forms and delaying the verification process for union member thresholds required for registration (Interview, 2019). The effect of these administrative attacks on unions – which gave employers ample opportunity to engage in union busting by dismissing labour activists before they had managed to register – was described by the Solidarity Center staffer as being akin to a bathtub without a plug. In his words, ‘For a time, the level of water in bathtub was increasing … now it's draining faster than the running water is coming in’ (Interview, January 2019).

The disputes resolution system also remained a point of contention during the period of NLD government, especially its failure to protect basic trade union rights. In 2018, the ILO reported persistent ‘anti-union discrimination, in particular dismissals and blacklisting of union leaders and fears amongst ordinary workers about dismissal if they join a union’ (ILO, 2018: 26). According to a labour representative on the National Arbitration Council, freedom of association and trade union rights were ‘worse’ under the NLD (Interview, January 2019). This interviewee reported that around 90% of cases before the Council pertained to dismissals of unionised workers and union leaders. A large proportion of these related to union registration, although ‘sackings occur even once you get a basic union registered’. Reflecting on the situation, he noted that ‘workers are not asking more than what the law allows but they are getting less than what the law allows’. While matters before the Council were, in his opinion, generally ‘handled well’, penalties for non-compliance remained so weak that Council rulings did not change employer behaviour.

Faced with these obstacles, unions continued to demand genuinely democratic reforms in the industrial relations system that would enable worker voice and participation. In response to the failures of the disputes resolution system, workers and union members again resorted to public protests to amplify their demands for reinstatement, including protest actions at government and NLD party offices. The government responded by using its powers available under the 2011 Right to Peaceful Assembly and Peaceful Procession Law and its associated guidance rules to limit union protest actions. The MICS General Secretary explained how he had been arrested, alongside leaders from other national federations, for participating in a labour demonstration in Mandalay based on an interpretation of the Law as meaning that only workers born in the city could participate in public protest actions (Interview, January 2020).

With space for union activity within Myanmar contracting, unions increasingly focused on increasing their engagement with international organisations and global companies – especially brand buyers in the garment manufacturing industry – around the need for further reform to address freedom of association violations and eroding trade union rights (Interview with CTUM Executive Member, 2020; Interviews with MICS Executive Members, 2019 and 2020). This task became even more urgent in the context of the COVID-19 pandemic, which led to further union busting through targeted dismissals of union members (Interviews with CTUM and MICS Officials, 2021). The government also used the pandemic as a justification for the suspension of the minimum wage determination process, one of the areas of labour governance in which unions had been able to secure clear gains for workers.

Unions continued to represent workers at the enterprise level, through public campaigns and via institutional engagement to try to protect and defend them from the worst impacts of the pandemic (Ford et al., 2023). They engaged with the NLD government and senior officials within the Ministry of Labour in an attempt to influence policies around factory closures, access to basic income in the event of lockdowns or quarantine, and safety protocols to protect workers and limit their exposure to the virus. They also engaged with their members, distributing checklists of COVID-19 preventive measures so their implementation could be observed in a specific workplace, and then taking up reported non-compliance with the Ministry of Labour (Interviews with CTUM Executive Member, 2021; MICS Executive Member, 2021).

Ultimately, the hostility of the state bureaucracy and the failure of the NLD government to support worker participation and representation prompted unionists to run labour-friendly political candidates in the 2020 national general election. These candidates ran in constituencies with a high concentration of workers and factories but were hampered in their ability to campaign by resource constraints and restrictions on public events during the pandemic. While none were successful, they nevertheless continued to believe that electoral participation was a necessary first step into linking democracy at work with democratic participation in elections in the face of perceived neglect by the NLD (Interview, CTUM official, December 2020).

A return to coercive control

After the coup of 1 February 2021, the military junta took strong action to assert control over all spheres of public life, using emergency powers to restrict freedoms and political opposition to their actions. As noted by the ILO, ‘the capacity of trade unions and CSOs has been severely curtailed by violent raids on their offices, the arrest and detention of key leaders, monitoring and restriction of finances and limiting access to communication networks’ (ILO, 2022: 4). Following the coup, the regime declared that sixteen unions and labour rights NGOs were ‘illegal’. Senior leaders went into hiding or were arrested because of their involvement in street demonstrations and mobilising resistance to the coup. The major national union federations still registered faced tight restrictions on their operations. Even factory-level union leaders feared that employers could call the police to make arrests and persecute a worker exercising any form of voice at the workplace. In short, as one union leader explained, unions’ capacity to represent workers was all but destroyed (Interview with CTUM Executive Member, 2021).

It proved difficult, however, to put the genie back in the bottle. As part of a broader social coalition, unions and labour NGOs stood at the forefront of street protests in the period immediately following the coup and have sustained their struggle against the junta in the face of violence and repression (Stokke, 2023). In the wake of the coup, almost all union federations withdrew from any form of institutional participation or cooperation with the regime. However, they continued to try to support workers by covert means or by communicating with supply chain actors such as brand buyers in the garment manufacturing industry about factory-level labour rights violations (Interview with CTUM Executive Member, September 2021).

Unions and union leaders based outside Myanmar or in areas of the country controlled by ethnic militias continued to operate, calling for comprehensive international trade sanctions and solidarity actions by the international labour movement to support the restoration of political democracy in Myanmar (Interview with CTUM Executive Member, November 2022). Unions also supported contestation of the credentials of regime officials to represent Myanmar in the ILO system (Myanmar Now, 2021) and the establishment of a special ILO Commission of Inquiry on the post-coup deterioration of labour and union rights. The Inquiry which found that there was ‘systematic and widespread’ use of forced labour by the military and a ‘persistent oppression’ of unions (ILO, 2023a: 10–11), has further served to delegitimise the regime government's tenuous claims to conforming with international standards and may be a precursor to formal sanction measures under the ILO constitution.

The utility and limits of authoritarian innovations

What does the coup of 2021, and the developments leading up to it, tell us about the utility and limits of authoritarian innovations, both in a broad, political sense and within labour governance? At a macro-political level, the institutional reforms that occurred in Myanmar across a range of governance domains between 2011 and 2021 constituted a political project of authoritarian innovation. These reforms did not originate from an immediate economic or political crisis, but rather from within the military political elite itself. They were not intended to promote wholesale democratisation of Myanmar's polity but rather to increase the country's economic prosperity and the domestic and international legitimacy of the state. The USDP government set out to achieve these aims through a performance of democratic reform – that brought with it many substantive improvements – but never a wholesale push towards democracy which it always envisaged as tempered by the ‘discipline’ of military authority.

Within this broader push, labour governance was a key domain of reform for a very clear reason. The government wanted to re-engage with the global economy, and at least nominal adherence to international labour standards was a prerequisite for doing so (Ford et al., 2017). While not seeking to fully integrate worker voice and representation in either the workplace or the policy arena, the initial wave of labour reform made it possible for workers to organise and register unions, which then (a) pursued workers’ interests through other newly established industrial relations processes for disputes resolution and wage-setting and (b) sought to build their power and influence in order to push for more space and more effective labour governance institutions. Even during this period, the state bureaucracy acted to constrain union power by changing protocols and regulatory requirements, allowing widespread union busting and failing to address the evident weaknesses in the other industrial relations institutions it had created.

These practices continued after the election of the NLD government. It is clear, moreover, that the NLD was wary of the consequences of strengthening of worker voice and democratic participation. As one of our interviewees suggested – with reference to both the state bureaucracy and the NLD government – ‘they want workers to be disciplined and not too many basic unions … they just want to have that control’ (Interview, 2020). Despite the effects of this desire for control over the unions, politicians and the bureaucracy failed to pacify workers and unions, which continued to push back against anti-worker measures. As a National Arbitration Council member observed, ‘If workers don’t have faith in labour officials at the township level, then they won’t have faith in the higher level’ (Interview, 2019). Illegal strikes and public protests were an inevitable corollary of bureaucratic hostility and institutional failure. As the MICS General Secretary concluded, ‘If workers believed in the ability of the law to address their problems, they wouldn't go to the minister's office’ (Interview, January 2020).

The union officials and workers we interviewed certainly conceived of themselves as democratic agents, essential to the emergence of a more genuine and ‘mature’ democracy. In the period before the coup, unions even looked towards moving into formal participation in election processes as a vehicle for challenging the hostile disposition of the state. Given more time, they could have followed in the footsteps of Indonesian unionists, who have increasingly inserted themselves in the political landscape in response to mainstream parties’ failure to support a labourist agenda (Caraway and Ford, 2020). Instead, their political ambitions were cut short by the coup, with which the hybrid experiment was abandoned, and replaced by a reversion to coercive authoritarianism. When powerful elites, especially in military-dominated states, revert to hard coercive authoritarian governance the disguised and incremental manoeuvres associated with authoritarian innovations are less necessary. And, the urgent task of democratic actors such as unions becomes one of direct resistance and struggle against the authoritarian character of the political regime.

Conclusion

In little more than a decade, Myanmar has moved through a pendulum swing from hard forms of militarised authoritarian governance to a political regime with hybrid characteristics, and then after the five-year term of a democratically elected national government, to overt authoritarianism. We have dissected the waves of institutional reform associated with these political shifts – and then their unwinding – on labour governance practice. In doing so, we have identified an initial expansion of space, albeit constrained, for labour voice and democratic participation that later again contracted as unions challenged a hostile bureaucracy and a democratically elected government that failed to recognise workers’ rights and representation as a legitimate space for consolidating democratic values. Across this 10-year period, we observed strong continuities in terms of how labour governance institutions and practices that appeared compatible or even supportive of workers’ democratic voice and participation actually restricted the space available to unions.

Importantly, however, workers and their organisations are not necessarily passive victims in such contexts. As part of our analysis, we have also drawn attention to the agency of an emerging union movement struggling in the face of the military's ongoing desire to retain influence and control over labour administration and the failure of the NLD government to recognise unions’ potential to contribute to the deepening and consolidation of democracy. And, as the Myanmar case reveals, even the narrowest windows for collective agency create opportunities for workers to push back against the constraints and intentions of authoritarian innovations in labour governance.

This analysis contributes to the literature on authoritarian innovations by examining novel meso-level labour governance practices that restricted union voice and capacity in the context of a country oscillating between ‘hybrid’ and authoritarian political regime types. It also points to the need for further studies of labour governance practices in countries controlled by hybrid regimes or experiencing some form of political transition. As the Myanmar case reveals, political transitions at the macro level – such as the installation of a democratically elected government – do not necessarily lead to sustained democratic reform of labour governance. As demonstrated here, a formal commitment to democracy may in fact be accompanied by authoritarian innovations in labour governance that serve to contain worker voice while maintaining the appearance of democracy, legitimacy and alignment with international labour standards.