New governance of labour rights: the perspective of Cambodian garment workers’ struggles

ABSTRACT

Globalization has implied the transfer of industrial work to countries of the Global South, where labour rights are seldom effectively protected by legal frameworks. New forms of governance that go beyond state-centred legal regulation are presented as an alternative to fill in ‘governance gaps’. This paper analyses ‘new governance’ from the perspective of Cambodian garment workers and labour movements’ struggles. Drawing on the literature on governance and private regulation and research data from Cambodia, it argues that a technocratic approach makes governance initiatives ignore the economic conflict between labour and capital but also the possible political conflict between labour and government. By ignoring trade union rights, power-blind initiatives might end up weakening both the labour movement and democratic accountability, instead of complementing state’s regulatory roles. This might serve the overlapping interests of the powerful actors both in Cambodia and internationally.

KEYWORDS:

• Labour rights
• trade union rights
• governance
• private regulation
• Cambodia

Neoliberal globalization has meant the transfer of industrial production to the countries of the Global South, where legal frameworks are often weak and/or their implementation insufficient to protect workers’ rights. Garment industry is an example of how production is organized into global supply chains, at the bottom of which work is precarious and labour rights violations widespread. These problems have been interpreted as consequences of ‘governance gaps’, which, it is argued, must be filled with new forms of regulation that complement and go beyond the traditional state-centred ones (Ruggie, 2014). ‘New governance’ is presented as a shift away from top-down models to more participatory settings, where regulatory functions are shared among stakeholders and binding law and treaties give way to softer, more flexible forms of regulation (Abbott & Snidal, 2009).

In the garment industry, this shift is reflected in the privatization of labour rights regulation and the growth of monitoring and audit regimes associated with the discourse of corporate social responsibility (CSR). However, a growing body of critical research has drawn attention to the weak enforcement of non-state governance and its tendency to overlook a particular category of labour rights, namely, trade union rights. This weakness is interesting from the perspective of labour studies, concerned with the ‘crisis of labour’ caused by globalization and capital’s mobility, that have weakened labour movement’s traditional sources of power and delegitimized it as a social actor (e.g. Silver, 2003).

This paper analyses the relationship between new governance and the crisis of labour through a case study examining the Cambodian garment industry. It analyses the diverse governance initiatives – aimed at complementing state’s regulatory roles – from the perspective of garment workers and labour movement’s strategies. Drawing on theoretical literature on new governance and private regulation of labour rights and research data from Cambodia, it illustrates how the power blindness of governance initiatives implies ignoring both the economic conflict inherent to the labour–capital relationship and the possible political conflict between organized labour and government. This suggests that technocratic approaches that weaken trade union rights – as has been shown in previous literature – not only undermine labour movements but also democratic accountability. In contexts like Cambodia, the assumption of complementarity between private and public governance seems rather like wishful thinking, because powerful actors benefit from the existence of governance gaps. The gaps are not coincidental, nor is the power blindness of governance initiatives expected to fill them: They are in line with the logics of global supply chains that build on the disempowerment of workers in the Global South (Phillips & Mieres, 2015; Taylor, 2009). Enforcement falls on foreign buyers, whose interest in enforcing trade union rights is very limited. Instead of filling gaps in state-centred regulation, governance initiatives might serve as a disguise under which governments can use their regulatory capacity for the opposite purpose, to limit trade union rights and claims for social justice.

The paper is structured as follows: The first part looks at how governance literature has presented the shift from old to new governance as a response to the adverse effects of globalization. Drawing on critical CSR research, labour studies, and sociolegal studies, the paper then suggests that the paradigm is analytically weak due to lacking attention to power relations and the economic conflict inherent to the labour–capital relationship. The second part focuses on Cambodia and presents the country’s pluralist labour rights governance setting, using mainly secondary sources. In the third part, this setting is analysed from the perspective of trade union rights and labour struggles. It is argued that the usefulness of governance initiatives for workers has been weakened by their technocratic approach that renders them vulnerable to the influence of the most powerful ‘stakeholders’, including the government.

The analysis presented in the third part is based on empirical data from interviews and observation among Cambodian garment workers and labour movement in Phnom Penh between January and March 2016 and between October 2016 and January 2017. The main data consist of 35 semi-structured interviews with two groups of informants. The first group consists of national-level union leaders and local level activists, labour NGO staff, as well representatives of international companies and governance initiatives. The second group includes ILO and other UN staff, journalists, and international NGO representatives that have a long history working on Cambodian labour issues and have been interviewed mainly as observers. The interviews form part of the data collected for a wider research project on Cambodian labour movement’s strategies since the 1990s. References to individual interviewees have been anonymized in order to protect their identity.

From old to new governance

The shift from old to new governance (or government to governance, Rosenau, 1992) of labour rights is often presented as a response to ‘governance gaps’ produced by globalization. Since the 1970s, northern-based multinationals began outsourcing their production offshore in order to reduce costs. In labour-intensive industries such as garment production, a major incentive was the difference in labour costs resulting from socio-economic and regulatory differences between the North and the South. Production was organized into global supply chains (e.g. Abbott & Snidal, 2009; Locke, 2013; Mayer & Gereffi, 2010) that enable Northern lead firms – garment brands and retailers – to avoid direct involvement in the labour-intensive phases of production. Workers who manufacture the products are employed by supplier companies.

As a result, production escaped beyond the reach of the old state-centered governance systems of the Northern countries. Governments of the new ‘host countries’ lacked the capacity to regulate powerful multinational companies and to enforce labour legislation – or the will to do so (Locke, 2013; Mayer & Gereffi, 2010). Deregulation and low labour costs were perceived as their main competitive advantage, according to the neoliberal doctrines of international financial institutions. The resulting situation was presented as a ‘governance deficit’, often described in Polanyian terms as a disembedding of market from institutions of governance (e.g. Marx, Wouters, & Rayp, 2014; Mayer & Gereffi, 2010; Ruggie, 2013).

Globalization of communications brought the effects of this ‘deficit’ – labour rights violations and abusive working conditions – to the attention of Northern publics, and vocal sweatshop campaigns lead to companies adopting codes of conduct (CoCs) and other CSR initiatives to protect their reputation (O’Rourke, 2003). At the inter-state level, the idea of creating a binding global regulatory system by including ‘social clauses’ into the WTO ended in failure (mainly due to fears of protectionism) and instead, international organizations began developing ‘soft’ initiatives, such as the UN Global Compact, to which companies could join directly, skipping over the state level (Locke, 2013). In governance literature, the emergence of new forms of regulation is seen as a sign of a Polanyian double movement, society’s reaction to the adverse effects of globalization and commodification, as the state ‘by itself cannot do all the heavy lifting required to meet most pressing societal challenges and … therefore needs others to leverage its capacities’ (Ruggie, 2014, p. 8).

The important differences between ‘new’ and ‘old’ governance (at national and transnational level) are the shift from ‘hard’ laws and treaties to ‘soft law’ approaches, and the transformation of state’s role. This does not mean that public regulation is simply replaced by private, but that state’s regulatory monopoly is opened to the participation of private actors such as companies and civil society, while state remains an ‘orchestrator’ (Abbott & Snidal, 2009).

New governance and labour rights

These two aspects of new governance are clearly reflected in labour rights regulation in global supply chains. For most of the twentieth century, labour standards were regulated by states through laws and a system of industrial relations and, at the international level, through the ILO’s system of labour conventions (Locke, 2013). Currently, multiple actors are involved in governing labour rights though initiatives ranging from auto-regulation and company CoCs, to multi-stakeholder initiatives (MSIs) and national and transnational private–public schemes such as Marocco’s Fibre Citoyenne or the ILO’s Better Work programme (see e.g. Anner, 2012; Locke, 2013; Mayer & Gereffi, 2010; O’Rourke, 2003; Seyfang, Pearson, & Jenkins, 2002; Wells, 2007).

These initiatives are legally non-binding: Their enforcement is based on market incentives and consumer pressure. In addition, most global governance initiatives (such as the UN Global Compact, UN Business and Human Rights Framework, OECD Guidelines, and MSIs) do not refer to the ILO’s legally binding international labour conventions, but to the 1998 Declaration of Fundamental Principles and Rights at Work that uses a softer language of ‘principles’ to define fundamental labour rights.¹ The ‘softness’ of new labour rights governance has been the object of most criticism in literature. Consumer pressure is relevant only for the most visible brands, and purchasing power is transferring to countries where consumers are less concerned about social issues (Barrientos, 2008; Merk, 2014). Most authors conclude that market-based initiatives are too weak to have an impact and must be complemented by state regulation to guarantee enforcement (Barrientos & Smith, 2007; Compa, 2008; Esbenshade, 2012; Mayer & Gereffi, 2010; Mezzadri, 2012; Prieto-Carrón, Lund-Thomsen, Chan, Muro, & Bhushan, 2006; Seidman, 2008; Wells, 2007). However, the question about whether – and through which mechanisms – public and private governance actually complement and strengthen each other, has begun to be examined only recently (see the Special Issue of Global Policy 8(S3), May 2017). The assumption of complementarity implies the existence of political will to protect labour rights. It must be carefully examined, particularly in the context of countries that compete against each other on the basis of low production costs.

Importance of trade union rights

Labour rights scholars have drawn attention to another weakness of new governance. Its positive impact tends to be limited to ‘outcome standards’ in occupational health and safety, but it fails to deliver results in terms of ‘enabling rights’ (Rodríquez-Garavito, 2005) or ‘process rights’ (Barrientos & Smith, 2007), particularly trade union rights (freedom of association and right to collective bargaining). Egels-Zandén and Merk (2014) describe how this happens in practise: buyers pay lip service to union rights, their violations are not detected in audits. Workers are treated as passive objects and parallel means of organizing are emphasized, and suppliers have limited incentives for compliance and violations are not remedied.

Barrientos and Smith (2007) explain this bias as a result of companies’ technocratic compliance approach. Labour rights monitoring is seen as just another audit – comparable to technical and financial ones – and easily measurable and quantifiable outcome standards become a natural focus. For Merry (2011), human rights audit culture is an example of the extension of corporate forms of thinking into the realm of the social. In this process, political debates about what constitutes human rights compliance are replaced by technical discussions about measurement methodologies. In the governance paradigm, public sphere tends to be presented as a ‘depoliticized arena of collaboration among generic “stakeholders”’ (Rodríquez-Garavito, 2005, p. 209) without considering how power imbalances affect the collaboration and its outcomes. This leads to inattentiveness to trade union rights.

Rodríquez-Garavito calls for the incorporation of empowering mechanisms into monitoring schemes. However, this is difficult if the failure of private regulation to consider and address power imbalances is not a coincidence but intended. The choice of indicators – what is measured as labour rights compliance – represents the perspective and power of those who define them (Merry, 2011), and as LeBaron and Lister (2015) point out, the politicized, productive power of CSR instruments is ‘intertwined with the expansion of corporate control and profitability’ (p. 907). Anner (2012) sees a logic behind the ‘corporate capture’ of regulation: Promotion of trade union rights weakens companies’ managerial control over the supply chain without producing much reputational value in comparison to issues such as child labour (see also Bartley & Egels-Zandén, 2015; Seidman, 2008; Utting, 2008).

Importantly, when rights are integrated into corporate governance and supply chain management, their nature as claims tends to be ignored and delegitimized (Scheper, 2015b). As Scheper notes, they become something to be implemented from above, in contradiction to Ingram’s (2008) suggestion that rights are only meaningful as an expression of autonomy and a result of contestation. While Ingram talks about human rights in general, this notion is even more fundamental in the case of labour rights. An economic conflict is inherently present in the relationship between worker and employer (Savage, 2008), or labour and capital. This relationship is fundamentally unequal in terms of power, which makes trade union rights central to all labour rights. Workers’ need to join forces and collectively defend their interests is the foundation of the trade union movement (Dunning, 1998) and the need to protect their freedom of association and right to form trade unions has been recognized in all the ILO’s fundamental documents since the organization’s 1919 Constitution. It forms one of the four fundamental principles of the 1998 Declaration. Novitz (2009) points out that while freedom of association is a general political right (and protected as such under the International Covenant on Civil and Political Rights), worker’s purpose in forming organizations is different from other types of actors, as it is meant to address long-standing inequalities in bargaining power. The logic is to protect the right to struggle, which is why Convention No. 87 on Freedom of Association is promoted together with the Convention No. 98 on the Right to Collective Bargaining. This is captured by Santos (2002) who considers the right to organize a metaright – a ‘right to create rights’.

In sum, the problem with the new governance of labour rights is not only related to the lack of binding norms, but to the way labour rights are (re)defined under these initiatives. When the conflict between ‘stakeholders’ is not recognized, interpreting labour rights from the perspective of the most powerful actors seems unproblematic. Trade union rights – needed to protect workers’ ‘right to create rights’ – can be ignored, without this appearing to be caused by a lack of (political) will.

Labour rights governance in Cambodia

Garment industry has had a central role in Cambodia’s social and economic transition since the 1990s. As the end of the Cold War changed the political balance in South-East Asia, Cambodia’s post-Khmer Rouge leadership turned to Western powers for financial and political support, needed for ending the civil war and gaining control over the country. To achieve this, the ruling party abandoned its communist ideology, adopted a new name (Cambodian People’s Party, CPP), and a discourse of democracy and human rights. Domestic support was secured through elaborate networks of patronage and corruption, while opposition voices were silenced with the threat of violence (Hughes, 2007; Strangio, 2014; Ward & Mouyly, 2016).

In the spirit of the 1990s donor consensus, democratization was to go hand in hand with liberalization of economy. Development was sought through export-led growth to which garment industry was an appealing motor due to low entry costs and a vast supply of cheap labour that helped attract foreign investors. The industry grew from practically a zero in 1994 to employing 80,000 workers by 1998, when negotiations to bring Cambodia under the United States garment import quota system were initiated. During the five-year US-Cambodia textile agreement (UCTA, 1999–2004) the sector grew to 220,000 workers (Polaski, 2006).

The first factories were opened by Asian investors. Currently, 33% of the ownership is Chinese, followed by Taiwan, Hong Kong, and South Korea. Cambodian ownership is only 3% (BFC, 2016). In 2015, garment and footwear industry accounted for 78% of the country’s total merchandise exports, with 699 licensed export factories employing around 650,000 workers (BFC, 2016). Despite the industry’s importance to the national economy, Cambodia is responsible for a mere 0.3% of global production which makes the industry very buyer-driven (Miller, Nuon, Aprill, & Certeza, 2007). Lack of industrial upgrading increases vulnerability: the sector still follows the 1990s model of cut-make-trim functions, imported inputs, and unskilled labour.

Cambodian labour rights regime² emerged from the 1990s democratization process and a comprehensive legislative reform undertaken to build the country’s international legitimacy (Ward & Mouyly, 2016). The legal framework on labour rights is very progressive. The 1993 Constitution guarantees the right to establish associations and to strike. Cambodia has ratified the main international labour rights instruments, including ILO Convention No.87 on Freedom of Association and Convention No. 98 on the Right to Collective Bargaining. They are reflected in the 1997 Labour Law, drafted with the support of the ILO and international donors, particularly the US. Despite the strong legal framework, implementation is weak. Labour inspection and the judicial system suffer from a lack of resources and omnipresent corruption.³ Labour courts provided in law have not been established and instead, international partners supported the creation of alternative governance institutions to complement state’s weak regulatory capacity.

ILO-lead monitoring programme

The main governance innovations originated from the UCTA, that innovatively linked improved labour rights compliance to import quota increases (contrary to previous trade agreements that had included punitive labour clauses) (Polaski, 2006). Decision-making regarding quotas required reliable information about compliance, and as state capacities were deemed insufficient, the ILO was requested to assume factory inspection (Kolben, 2004; Polaski, 2006; Shea, Nakayama, & Heymann, 2010). The monitoring project started in 2001. Compliance with the national and international law was measured with a 500-item checklist during bi-annual factory visits. Findings were to be reported to each factory and optionally buyers. In addition, public reports would provide information on overall developments in the industry, naming individual factories if they failed to remedy problems.

Early observers considered the project a success story. They claimed it created an unprecedented amount of data and provided multiple compliance incentives, leading to employment creation, improved labour rights, and economic growth (Polaski, 2006; see also Kolben, 2004; Sibbel & Borrmann, 2007; Wells, 2006). At the end of UCTA’s in 2004, international lobby and the Cambodian government’s official request resulted in the extension of the project. Renamed Better Factories Cambodia (BFC), the project was to be transferred to Cambodian authorities by 2009, but its mandate has been continuously renewed.⁴

At this point, the project’s leverage structure changed significantly. Between 2004 and 2014, BFC did not publish information on individual factories and while BFC membership became a requirement for an export permit, non-compliance has never lead to revoking it (Human Rights Watch [HRW], 2015), which means that leverage fell exclusively to buyers. However, only 13 buyers were accessing BFC reports for sourcing decision purposes in 2006 (Miller et al., 2007) and still in 2015, 37% of the factories had no buyers purchasing the reports (BFC, 2016).

In the following years, several critical NGO studies claimed that BFC reports were providing too positive a picture of labour rights in Cambodia, where garment workers’ real wages had decreased by 22% between 2001 and 2011 (Workers Rights Consortium [WRC], 2013). They suggested that lack of transparency was undermining the programme’s impact (HRW, 2015; Merk, 2012; Sonnenberg & Hensler, 2013). Academic scholars’ enthusiasm also decreased, pointing to moderately positive results in minimum legal requirements particularly related to occupational health and safety (Berik & Rodgers, 2010; Merk, 2012; Miller et al., 2007; Shea et al., 2010). In interviews, BFC representatives explained that the internal recognition of these weaknesses led, in 2014, to the launch of a ‘Transparency Database’, a tool for publishing information on individual factories compliance. Together with increasing advisory services for factories, it has improved compliance (interview, BFC1).

Innovative arbitration of disputes

An indirect result of the UCTA was the Labour Arbitration Council (AC), foreseen in the Labour Law but created in 2003 under an ILO dispute resolution project funded by the US (De Noord, Hwang, & Bugeja, 2011). It is mandated to resolve disputes between workers and employers that are forwarded to it by the Ministry of Labour. A tripartite arbitration panel – chosen by the parties to the dispute from a 30-member list of arbitrators – must organize a hearing and emit an arbitration award within 15 days (Adler, Sage, & Woolcock, 2009; De Noord et al., 2011).

Initially, the arbitrators got legal and administrative support from the ILO and a local labour rights NGO. In 2005 these tasks and fundraising were transferred to Arbitration Council Foundation (ACF). The number of cases grew steadily until 2015, when AC handled 338 cases (260 in garment/footwear) (BFC, 2016), mainly related to wages and bonuses, and secondly to discipline and termination of employment. By the end of 2015, it had dealt with over 2400 cases, affecting around one million workers and employers. It is currently funded by the Swedish cooperation SIDA and USAID (interview, AC1).

The UCTA provided the US important leverage that it used to support the ILO in pushing for a carefully designed structure of checks and balances to guarantee the independence of AC (interview, UN1). The arbitrators were carefully chosen and cannot be easily dismissed, which protects them from political influence. Contrary to the judicial system, AC is widely perceived as neutral and non-corrupt (Shea et al., 2010) and has a central role in labour rights protection in Cambodia. Its awards are considered authoritative juridical opinions and contrary to court decisions, they are published, which makes AC the first institution to build jurisprudence on labour law in Cambodia.

The awards are binding only if both parties agree to it. According to ACF director, this protects the institution from corruption, as it makes bribes useless (interview AC1). The quality of awards is the basis of ACs authority and recognized by buyers, whom workers usually turn to – instead of courts – if an employer does not comply with the award. In 2010 the ILO facilitated the signing of a Memorandum of Understanding between the employer’s association (Garment Manufacturers Association in Cambodia, GMAC) and unions. According to the MoU, companies would treat all awards as binding, in response to the union side respecting AC’s return-to-work-orders. It expired in 2014, and GMAC has refused to sign a new one.

AC is currently under considerable change. The new trade union law of May 2016 has limited independent unions’ access to AC, and the number of hearings dropped by 42% in the second half of 2016 compared to the same period in 2015 (data available on AC website). This law will be discussed in more detail in the following chapter.

Buyer-lead governance

Garment factories in Cambodia are mostly owned by Asian supplier companies that sell to European and North-American buyers. There are big differences between the practices of different companies, which Oka (2010) associates to different levels of reputation-consciousness and vulnerability to publicity. This chapter describes some general tendencies.

Generally, buyers include CoCs and requirements of third-party certifications as binding obligations in their sourcing contracts with suppliers. Before orders can be placed, a new factory must pass an audit that is based on the buyer’s corporate compliance checklist and conducted by internal sustainability staff (or an external social auditing company). During the business relationship, continuous monitoring uses the same or similar checklist and/or BFC reports. Some buyers do not source directly from factories but contract intermediary buying companies, who are in this case responsible also for ‘social compliance’.

While licensed export factories are regularly monitored, the challenges that have been widely acknowledged in literature (see LeBaron and Lister’s synthesis (2015)) are also present in Cambodia. The workers the author has talked to, do not believe that any workers tell the truth about working conditions to auditors. They are too afraid of the management finding out. Moreover, as LeBaron and Lister (2015) note, social audits’ capacity to detect and report non-compliance is shaped by their corporate client. The contents, indicators, and requested level of detail varies significantly between the different CoCs and certificates. According to GMAC, even the most visible brands do not care about how well their suppliers actually score in audits, as long as they are not qualified as non-compliant, i.e. in violation of the most serious issues, such as child labour. Improvements beyond the minimum level are not rewarded by buyers (interviews C1 and C3; see also Oka, 2010) and incentive for undertaking more demanding ‘extra’ audits to attract buyers is practically inexistent for suppliers, who bear audit costs. Only two factories in the country have been certified by SA8000, the strictest certificate in terms of social compliance (interview C4).

If important or persistent non-compliance is detected, buyers demand corrective action plans. As the sustainability manager of a large European brand described, the approach is one of ‘continuous collaboration and dialogue’, aimed at ‘resolving challenges together’ (interview C2). Cancelling orders or terminating the relationship is the last option. This is in accordance with worker representatives’ wish, as they are ultimately afraid of workers losing their jobs.

The largest reputation-conscious buyers like H&M, GAP, or Adidas have staff in-country, and they participate actively in BFC and might fund labour rights-related projects. However, unionists complain that even ‘the best’ buyers’ interest in talking about labour rights is not reflected in their actual practises. Buyers approach to labour rights is technical, and they aim at keeping their responsibility narrow. In the end, they are neither employers nor law-makers, as a sustainability manager interviewed by the author pointed out (interview C2).

Cambodian governance and labour’s struggles

DiCaprio (2013) suggests that BFC has guaranteed trade union rights in Cambodia and AC has created a space for exercising these rights. The combined effect is worker empowerment that ensures the sustainability of the Cambodian labour rights regime. A more critical view is presented here, arguing that the initiatives’ impact and sustainability has been limited by their vulnerability to pressures by the most powerful economic and political actors.

Trade union rights in Cambodia

When the first independent unions⁵ were emerging from garment workers’ grassroots struggles in the 1990s (Hughes, 2007), they faced difficulties in even getting registered (Kolben, 2004; Ward & Mouyly, 2016). However, union numbers increased rapidly from 20 in 1997 to around 1000 in 2006 (Shea et al., 2010) and current estimations of total union numbers are around several thousands. As of October 2016, 104 national trade union federations (and 13 professional associations), grouped in 13 confederations, were registered with the Ministry of Labour. Eighty-nine of them were active in the garment sector (American Center for International Labor Solidarity [ACILS], 2016).

While some authors interpret this impressive growth as a sign of union empowerment (DiCaprio, 2013; Oka, 2016) others see it as fragmentation that weakens independent unions and their political activism (see Arnold, 2013; Arnold & Han Shih, 2010; Hughes, 2007; Sonnenberg & Hensler, 2013; Ward & Mouyly, 2016). Both labour activist and observer interviewees believe it is a deliberate strategy by the Government and employers. For the most part, new unions have emerged from internal divisions of government- or employer-controlled unions (Arnold & Han Shih, 2010) or belong to ‘mafia unions’ (Arnold, 2013) dedicated to extorting money through strikes. All except three confederations and a handful of non-affiliated federations are officially aligned to the ruling party (ACILS, 2016). They form patronage networks under high-ranking government officials. According to an ILO observer, the purpose of these networks is more political than economic, and they are activated for electoral purposes (interview UN1). Company unions usually belong to CPP-aligned confederations. Cambodian interviewees often use the term ‘yellow union’ also for government unions, which indicates somewhat blurry boundaries between the two.

The dominance of government- or employer-controlled unions arises in the interviews as one of the main obstacles for the defence of garment workers’ rights (in line with observations by Arnold and Han Shih (2010) and Ward and Mouyly (2016)). It hinders worker representation and collective bargaining, because law requires the employer to negotiate only with unions with ‘most representative status’ (MRS) – often ‘yellow’ unions to which management might add new workers automatically. Independent unions have practically no collective agreements in the sector. The situation also weakens workers’ bargaining position, as most unions take employer or government side in negotiations and rarely participate in national demonstrations (Oka, 2016) (which probably contributed to the wage stasis of the 2000s; see Arnold & Han Shih, 2010; Hughes, 2007). Importantly, the dominance of profit-seeking unions contributes to an overall delegitimization of the labour movement and enables Government and employers to label all strikes as extortion.

Ward and Mouyly (2016) describe a collusion between the government and employers who share the common goal of avoiding independent unionization (also Arnold, 2013; Arnold & Han Shih, 2010; Hughes, 2007; HRW, 2015). Employers want to avoid demands for higher wages and to prevent strikes and penalties that buyers impose for delays in the on-time industry (Oka, 2010). Interviewees tell about employers’ aggressive attitude against independent organizing, and point to the increasing use of short-term contracts. They are used to put workers under constant threat of losing their jobs, in order to prevent organizing (see also Arnold & Han Shih, 2010; HRW, 2015; Sonnenberg & Hensler, 2013; Yale Law School, 2011).

Fears of union power grew when a national strike was followed by unprecedented wage increases in 2014. Spontaneous protests started in December 2013 when the government rejected a $160 minimum wage suggested by a government-set tripartite body, and announced instead a meagre rise from $80 to $95 for 2014. In the first days of 2014, protests grew into a national-level strike to which employer side reacted with factory close-outs and threats about moving production out of Cambodia. Government response, a violent crackdown by military police, lead to the death of four workers and around 30 wounded (AMRC et al., 2014; WRC, 2014), which raised significant international attention. The labour movement used this to build a successful minimum wage campaign that was supported by an international coalition and several buyers, and the Government agreed to a $128 minimum wage for 2015 (subsequently $140 for 2016 and $153 for 2017).

For a large part, the increases merely compensated for the decreasing real wages of the previous decades. However, they were nominally impressive and a remarkable show of labour’s increasing structural and associational – or economic and political – power. An equally remarkable response by employers and the Government followed. Criminal charges were raised against protest leaders based on GMAC complaints and registration of independent unions was de facto suspended. A draft law to regulate trade unions, that GMAC had been pushing for since 2008 (interview C1),⁶ was revived. Passed in May 2016, the law poses serious limits to freedom of association. It defines complex criteria for union organization and registration as well as strikes, and limits access to dispute resolution. It permits dissolving a union on the grounds of ‘serious misconduct’ or offense committed by an individual leader, which can be used together with the criminal code to silence unions (interview UN2). The 2014 protest leaders are under judicial supervision and their criminal charges are kept pending, which several interviewees consider a reason for decreasing public protests. A UN observer reflected: ‘In terms of speaking in public, you can see a big change. Trade union leaders, they are afraid compared to before. Before the trade union law they were brave … And now they have the criminal charges’ (interview UN2).

Ward and Mouyly (2016) claim that GMAC has succeeded in pressuring the government to support companies’ interests. However, maintaining a divided and co-opted labour movement is also in the Government’s own interest. High-level officials’ and CPP elite’s economic involvement in factories and special economic zones⁷ is intertwined with political patronage. Independent unions are also seen as a threat due to their historical links with the political opposition. The national strike occurred in the context of months-long protests that followed the 2013 national elections. In addition to losing popularity for the first time since 1997, the ruling party was accused of fraud. In this context, the labour movement and the opposition party CNRP (Cambodia National Rescue Party) saw each other as strategic allies, and CNRP actively mobilized workers to join protests. The possible fusion of the two protest movements is commonly seen as the ultimate reason behind the violent crackdown of the strike.

Relevance of governance initiatives to labour struggles

In her account of how the international development intervention undermined political contestation in Cambodia in the early 2000s, Hughes (2007) claims that BFC weakened the labour movement by capturing militant leaders into disempowering negotiation structures and delegitimizing grassroots struggles. Giving each union an equal seat in negotiation tables favoured the numerically dominant pro-government group and legitimized them as worker representatives. The technocratic monitoring regime avoided political framing of labour rights violations, and instead focused on issues like child labour.

The ILO project that preceded BFC protected the emerging labour movement by requiring respect of labour law, including the right to organize and register unions. However, BFCs role in strengthening trade union rights and supporting independent unions’ struggles seems less clear. While BFC collects detailed information on freedom of association in the factories, it is not reflected in the public reports’ aggregate checklist approach nor the transparency database. Factory-level information is not public nor accessible to unions. The same applies to the efforts to tackle the question of yellow unions under the advisory services (described to the author by BFC representatives) and to BFC's advocacy with the Government and occasional conciliation of factory-level disputes. Because this important work happens behind the scenes, it does not provide public support to independent unions’ struggles or to delegitimizing union busting. Most union representatives are not aware of these efforts nor their effects, and seem to consider the programme a lost cause for their struggles. ‘I think mostly it’s just to clean the face of Cambodian export. Clean small part of the dirt, but don’t touch the real root causes. You cannot touch the real root causes’ (interview U2).

BFC’s careful and depoliticized approach, criticized by Hughes and the unionists, is to a large extent a reflection of the ILO’s position as a tripartite intergovernmental organization. It cannot get involved with political issues within the three sectors unless there is a clear violation of an international convention. Moreover, BFC’s position as a voluntary monitoring programme, that sells services to companies, is weaker than ILO’s legal role in monitoring international labour law. When UCTA ended, international interest in continuing the success story was high. It would seem likely that the Government and GMAC were able to use it to negotiate a more favourable version of the programme. Since 2004, BFC’s existence depended on them, and its’ previously more vocal role was replaced by a technocratic approach (Sonnenberg & Hensler, 2013).

When talking about their strategies, Cambodian labour activists do not bring up BFC unless asked about it, but they recurrently refer to AC. As described by one NGO staffer, when workers’ legal rights have been violated, unions first try to negotiate with factory owners (often based outside of Cambodia). This seldom succeeds, and local leaders take the case to the Ministry of Labour from where it is eventually sent to AC. If the employer does not respect the award, unions recur to buyers instead of going to corrupt courts. Here, awards provide concrete juridical arguments and legitimacy for the claims. According to both labour and company interviewees, reputation-conscious buyers require suppliers to respect AC awards, and have in some cases directly told their supplier not to go to court after a non-favourable AC award (interview L2).

The different weight of BFC and AC in labour strategies is explained by their different mandates, institutional designs, and embeddedness, i.e. their vulnerability to external pressure. AC’s structure as an independent institution established in law puts it in a stronger position vis-à-vis the Ministry of Labour and employers. Protected from political and economic pressures, AC can confirm trade union rights even against the Ministry’s interpretation, for example in the case of fixed-term contracts, which AC concerns unlawful. Adler et al. (2009) describe the AC as a ‘non-threatening yet transformative’ space of contestation that enables opening issues of power to negotiation. However, recent developments suggest that this might have been too optimistic. The new trade union law – Government’s reaction to the independent unions’ increased power and the 2013–2014 national strike – seems to have radically changed the power balance. GMAC refused to sign a new MoU on binding awards, because it would not be needed when strikes would be prevented by law (interview C1). More importantly, the new law defines the right to represent workers in collective disputes as an exclusive right of MRS unions. After the law was passed, the Ministry of Labour stopped forwarding minority unions’ cases to the AC. If this practise continues, it closes the doors of AC to minority unions. This would have an important impact on union busting cases, such as the dismissal of workers who try to form an independent union, where AC was the only institution that provided protection. It is too soon to assess the law’s long-term impact, but it is clearly a sign of the Government’s will and ability to propose and enforce legislation aimed at limiting trade union rights and undermining AC’s transformative impact on power relations. Moreover, in late 2016, it was followed by the zero draft of a law on labour dispute resolution process seemingly aimed at bringing the institution under the Ministry’s control (interviews L4, UN1).

Positive evaluations of Cambodian labour rights governance have emphasized the complementarity between external monitoring and state enforcement (Polaski, 2006; Wetterberg, 2011). BFC and AC are mandated to monitor and resolve disputes, but enforcement is to be undertaken by state institutions. Labour inspection should follow up on BFC, and courts enforce AC’s binding awards, but in both cases, enforcement falls instead on international buyers and becomes part of their CSR.

Buyers do have power vis-vis their suppliers, who are bound by contractual obligations; and the government, due to the sector’s importance for the national economy. As one NGO worker put it: ‘If there is no investment from the buyers, it all goes down’ (interview L1). Compared to local elites and Chinese investors, buyers – at least the reputation-conscious ones – are also seen as the only actor interested in rights. This makes a big difference between union struggles in the garment industry and other sectors:

It’s special, because in the garments you have brands, you have international relations (…) The level of respect of human rights between international company and national company is different: They can be hurt, like for instance if one of the H&M companies kills workers, it’s the end of [its] life, right? (interview, U2)

Several big buyers participate in industrial relations projects funded by themselves or external donors such as the German cooperation agency. However, the focus of the projects is usually not on trade union rights, but on preventing strikes – extremely common and used by some unions for money-making purposes – through improved communications. H&M and Inditex have signed global framework agreements with IndustriALL and H&M was about to establish a bipartite monitoring committee to coordinate its implementation in Cambodia in 2017.

Buyers’ increasing role in labour rights enforcement is reflected in union strategies. Independent union leaders and NGO staff are very knowledgeable about different buyers’ behaviour and reputation-consciousness. They actively approach ‘good’ buyers for support in factory-level conflicts, either directly or through international networks, particularly IndustriAll Global Union, or NGOs such as the Clean Clothes Campaign or Workers’ Rights Consortium.

However, labour representatives also see certain hypocrisy in buyers’ pro-labour discourse, and have doubts about buyers’ ultimate goals. One union leader used the term ‘deadly dialogue’, to describe the buyers’ willingness to meet and ‘talk and talk and talk’, without really changing action (interview U2). An NGO staffer commented:

They have their own agendas. For the brands, the goal is to secure their popularity. They don’t mind, also in Bangladesh (after the collapse of Rana Plaza) they just continue to order! What does it mean? They say they want to respect freedom of association but they buy from China, they buy from Vietnam! They only care if it affects their reputation. (interview L4)

Labour and employer interviewees both claim that at the end, buyers only care about price: They could guarantee good working conditions by paying better margins and avoiding impossible delivery terms that – as buyers know – are impossible to comply with without excessive overtime or subcontracting. However, buyer representatives explain that setting prices above the industry level is not possible if competitors do not do the same. For one international labour NGO this is what typically happens: ‘They just talk you around and claim that they can’t do anything more!’ (interview L5).

Even the most ‘reputation conscious’ buyers approach tends to be technical. They avoid talking about the economic conflict between workers, employers, and buyers, or the political nature of trade union rights. Buyers do not sufficiently examine whether unions with their Cambodian suppliers represent worker concerns (HRW, 2015). Buyer representatives explained to the author that while clearly management-controlled unions would be considered a violation of their CoCs, they cannot tell suppliers what kind of unions workers should have. One (foreign) sustainability representative suggested that the problem of union representativeness is something that the Government could take care of (interview C2) which indicates a weak understanding of the problem in Cambodia. Another example is the Fair Living Wage programme of H&M, considered a highly promising buyer-led initiative that is ‘based on negotiating wages between suppliers and democratically elected worker representatives’ (H&M, 2016). However, a labour NGO representative explained that their union partners have not been able to identify the Cambodian pilot factory, which could suggest that the initiative is piloted in a factory with only government-controlled unions. While this might be a technical detail to the buyer, it is in practise an effective way of legitimizing non-independent unions and weakening independent organization.

In Cambodia, the implications of the technocratic approach are wide-reaching, as all the independent mobilization is aggressively opposed by the Government. A depoliticized approach to labour rights disempowers independent unions. It undermines labour’s capacity to struggle and make economic and political claims, and delegitimizes labour’s protest as a form of democratic participation. Defence of labour rights shifts to Northern consumers and enforcement to Northern buyers – and responsibility of the Cambodian state to the background. Power-blind governance initiatives enable the state to maintain international legitimacy (and buyers’ orders), while preserving elite’s economic interests and political stability (Ward & Mouyly, 2016). The risk is that instead of complementing the state’s role, technocratic governance initiatives provide a disguise behind which the state can use its regulatory power – and capacity – in the opposite direction, to restrict trade union rights and limit spaces that might transform power relations, such as the AC.

Concluding reflections

The paper has examined new governance from the perspective of workers’ struggles to build and defend their labour rights, arguing that the positive impact of governance initiatives depends on their ability to take into account and modify unequal power relations. Moreover, it has shown that by ignoring trade union rights, governance initiatives risk weakening not only the labour movement but also democratic accountability.

The paper examined findings from previous literature in the light of empirical realities in the Cambodian garment industry, where diverse governance initiatives are implemented to complement state’s regulatory role. The ILO-lead labour rights monitoring programme BFC collects detailed data about export factories, but does not publish it. After losing the leverage of UCTA, BFC has to appear ‘neutral’ in the eyes of the powerful actors, government, and employers. The depoliticized tone of BFC’s public reports parallels the technocratic approach of buyer-lead initiatives that present labour rights as outcome standards to be audited with checklists, ignoring underlying power imbalances. AC has been better protected from external pressures, due to its institutional design and legal embeddedness. Its awards have provided legitimacy and practical tools for workers’ struggles. Even though lack of enforcement has weakened AC’s potential to modify existing power relations, its awards have been successfully used by Cambodian activists and their international allies to target buyers, who are de facto responsible for labour rights enforcement in Cambodia.

Instead of complementing the role of the state, governance initiatives seem to be substituting it, as buyers are increasingly given roles in enforcement of law.⁸ They fulfil these roles according to their commercial interests and CSR policies, presenting rights as outcomes, separate from issues of power and conflict. This is an example of how rights are constructed as a common interest in a process of hegemonization that reflects buyers’ interests (Scheper, 2015a; also Utting, 2008). While corporate CoCs might be binding, companies’ incentives to enforce trade union rights are limited by capitalist competition and the logics of supply chains that draw on a global division of labour based on the disempowerment of workers in the Global South (Phillips & Mieres, 2015; Taylor, 2009).

When the conflict between ‘stakeholders’ is not recognized, interpreting labour rights from the perspective of the most powerful actors becomes unproblematic, and trade union rights lose their original meaning as protection of workers right to ‘create rights’. Depoliticized governance approaches undermine workers’ possibilities to struggle and to claim rights from employers and the government, which in turn undermines democratic accountability, reducing the options to build political pressure in favour of pro-labour public policy and regulation.

In the case of Cambodia, the interests of global elites – international brands and retailers – seem to coincide with those of the local ones – Government and employers. All want the reputation benefit of appearing to support labour rights, without paying the costs of facing a strong independent labour movement. The power-blind governance approach ignores the conflict between labour and capital, but also the possible political one between organized labour and government. The Cambodian case suggests that it might be analytically beneficial to examine ‘governance gaps’ not only as a problem of regulatory capacity or weakness, but also of political will and economic interest. This bears similarity to the observation that Baker and Milne (2015) make about instrumentalism behind what is conventionally considered as ‘weak fiscal capacity’ of South-East Asian regimes: revenues from illicit economies often provide greater economic and political benefits than their regulation and taxation.

Acknowledging the powerful interests behind the ‘gaps’ enables us to better understand the challenges of strengthening trade union rights, particularly in the Global South. The analysis of private regulation might benefit from more attention to the question of political will, and the role of the state as ‘orchestrator’ of new governance, particularly in undemocratic contexts under current global capitalism (see Chan & Hui, 2016). Anner (2017) has recently written about the weakness of social compliance to support freedom of association, pointing to a need of state reform, building its capacity and legitimacy. An important question that remains is whether technocratic private regulation is too simply weak for its own job, or actually contributes to weakening democracy by depoliticising labour rights and labour movements. The Cambodian case suggests that this might be the case. Governance initiatives seem to form a disguise behind which the state uses its regulatory capacity to limit trade union rights and the transformative impact of spaces such as AC. Thus, the assumption of complementarity between public and private governance in contexts where governments are not accountable to citizens, seem feeble.

Interview data

U1, union leader, January 2017

U2, union leader, March 2016

U3, union leader, January 2017

L1, labour NGO worker, March 2016

L2, labour NGO lawyer, November 2016

L3, international labour NGO representative, December 2016

L4, labour NGO worker, December 2016

L5, international labour NGO representative, December 2016

C1, GMAC secretary, December 2016

C2, buyer sustainability manager, March 2016

C3, buyer sustainability staff, January 2017

C4, auditing company manager, February 2016

BFC1, two BFC representatives, March 2016

AC1, ACF executive director, March 2016

UN1, ILO worker, December 2016

UN2, UN worker, November 2016

Acknowledgements

The author wants to thank Globalizations editor Kevin Gray and two anonymous referees whose comments and suggestions significantly strengthened the article. She is also thankful for the comments made on previous drafts of the paper by Barry Gills, Anja Nygren and Christian Scheper, as well as the members of the Panel on Sociology of Law at the 28th Conference of the Nordic Sociological Association, August 2016. Finally, the author is deeply grateful to all the people in Cambodia who shared their time, observations and experiences with her.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The author would like to thank Kone Foundation for funding the research.

Notes on contributors

Anna Salmivaara

Anna Salmivaara is a doctoral candidate in development studies at the Department of Political and Economic Studies, University of Helsinki. She is currently working on her dissertation on labour rights regulation and labour movements in Cambodia, with a particular interest in the meanings and uses given to labour rights, law and regulation in different actors’ strategies. Her wider research interests include the relationship between neoliberalism and democracy, and the role of law and rights within this relationship.

Notes

1. While some legal scholars have greeted the Declaration as a sign of legal reflexivity at the ILO (Rogowski, 2015), authors like Alston (2004) consider it to constitute a revolutionary transformation in labour rights promotion, a shift from enforceable standards to soft promotionalism that opens the door to new interpretations of labour rights and risks destroying the ILO’s monitoring regime (see also Novitz, 2009).

2. “Cambodian regime” is used here to refer to garment export sector, the only sector covered by these governance initiatives.

3. Cambodia was ranked as 150 among 168 countries in Transparency International’s 2015 Corruption Perception Index.

4. The model has been expanded into Better Work, a global program implemented in 10 countries by the ILO and International Finance Corporation.

5. The term ‘independent union’ is used in the article to refer to unions that do not belong to confederations that are officially aligned to the ruling party, nor ‘yellow unions’ to which workers are added automatically upon recruitment. It is not used as a normative category, nor to claim that they have no ties to other actors than their members.

6. The ILO Committee of Experts had also called the Ministry of Labour to draft a law on trade unions in order to strengthen the protection of freedom of association (interview with an ILO representative).

7. Numerous interviewees talk about economic ties (ownership or corruption ties) between companies and high-ranking officials as common knowledge, a sine qua non for operating a business in Cambodia successfully. Individual cases have been uncovered by NGOs and media, but official information is not available.

8. Similar tendency has been identified in other countries, see for example, Berik (2017).