Contested compliance regimes in global production networks: Insights from the Bangladesh garment industry

Abstract

This article reports the findings of a field study on the emergence of collective agreements led by global brands enacting compliance measures to improve safety and working conditions in the Bangladesh garment industry. We explore how key actors in the Bangladesh garment sector who constitute the local production system of the global supply chain experienced the implementation of global agreements on factory safety. We argue that global safety compliance measures through multi-stakeholder initiatives provide legitimacy to multinational corporations and their global brands but do little to address the structural problems arising from exploitative pricing and procurement practices, which are the key reasons for deplorable working conditions in garment factories. Our findings indicate that neoliberal development policies of the state, where local economies are incorporated into global production networks, resulted in differential treatment and regulation of specific populations that comprise garment factory workers. The reconfiguration of state power to meet the demands of global supply chains also involved use of state violence to suppress dissent while undermining labour rights and working conditions. Our article contributes to the politicization of multinational corporations in global production chains by showing how contestations between workers, factory owners, the state, trade unions and multinational corporations create new private forms of governance and new regimes of compliance in the industry.

Keywords

corporate social responsibility garment industry global compliance regimes global supply chains private governance

Introduction

The collapse of Rana Plaza Tower in Bangladesh in 2013, leading to the deaths of 1133 garment factory workers, is one of the worst industrial disasters in recent times. The accident brought worldwide attention to the plight of workers in the readymade garment (RMG) sector in Bangladesh who are part of the global supply chain in a global production system where big name retailers and brands in the global north use cheap labour from poor countries in the global south to produce garments. These global production networks appear to be a win-win situation for all actors: consumers in Europe and North America are able to purchase cheap clothes, the retailers and apparel corporations increase their revenues and profits while people living in poverty in developing countries, especially women, are offered new opportunities to make a living by working in the factories that supply the garments. The reality, as evidenced by the Rana Plaza disaster, is somewhat grimmer. Extremely low wages owing to cut-rate pricing policies of buyers, unsafe working conditions, overtime working hours often without payment and a lack of collective workers’ rights are the norms in this sector (Anwary, 2017). The Rana Plaza disaster also called into question corporate social responsibility (CSR) practices and ethical buying commitments of global retail brands in the garment sector.

Following the accident, two multi-stakeholder initiatives were implemented with the aim of improving worker safety in the sector: The Accord for Fire and Building Safety (hereinafter referred to as the Accord), and the Alliance for Bangladesh Worker Safety (hereinafter referred to as the Alliance). The Accord and Alliance are competing private governance mechanisms of the global supply chain: both claim to be ‘legally binding’ to different degrees. The Accord claims to have more worker representation, whereas the Alliance is more a CSR-based approach, generally favoured by US corporations (Donaghey and Reinecke, 2017; van der Heijden and Zandvliet, 2014).

In this article, we explore how key actors in the Bangladesh garment sector who constitute the local production system of the global supply chain experienced the implementation of global agreements on factory safety compliance. In particular, we examine interactions between the state, local factory owners, workers, trade unions and non-governmental organizations (NGOs) and how these interactions reconfigure norms of governance by creating new regimes of compliance. How are global compliance measures enacted on the ground? What are the experiences of local actors in the global supply chain in implementing these compliance measures? To what extent do the compliance agreements and codes of conduct address the material needs of workers and ensure worker rights? These are some of the questions we will explore in this article. Our article makes three contributions: first, it enhances our theoretical understanding of global production networks by analysing the politics that define market, state and civil society dynamics. In particular, our findings reveal the types of class alignment and fragmentations that are being created in developing a labour force to serve global supply chains. Second, our article makes an empirical contribution by documenting lived experiences of individual actors embedded in global production networks. Third, our article provides a critical analysis of compliance and governance regimes that have emerged in order to manage global production networks. The politicization of multinational corporations (MNCs) is reflected by their power and capability to develop ‘inclusive’ but private regimes of governance involving market, state and civil society actors, and our article highlights the limits of these regimes in addressing oppressive labour conditions.

Global production networks, graduated sovereignty and labour rights

The global rise of economic neoliberalism has resulted in significant shifts in relationships between the market, state and society as well as in the international division of labour. Structural adjustment programmes imposed by the World Bank on developing countries compelled states to create investment and business friendly climates. As Harvey (2005) points out, state policies to encourage foreign investment are structurally biased against labour because wages need to be kept low to ensure competitiveness. Moreover, ‘flexible’ labour markets and weak environmental regulation made it attractive for corporations from the global north to source their products from developing countries given the high wages and rising costs of environmental compliance in their home countries. When social and environmental welfare of populations are in conflict with national competitiveness, the neoliberal developmental state tends to favour economic interests. In addition, state coercive power is used to repress dissent against economic reforms that favour corporate interest.

Global brands and retailers exert more control over production networks and supply chains than local manufacturers in the garment industry. Global production networks are ‘integrated economic, political, and discursive systems in which market and political power are intertwined’ and are driven by the logic of global capital (Levy, 2008: 943). These networks are characterized by power dynamics and contestations between the various actors in the supply chain. At one level, global supply chains are seen as driving forces of ‘development’ and poverty reduction in developing countries as they generate employment for local workers and contribute to economic growth. However, there are vast disparities in working conditions in this international division of labour: extremely low wages, unsafe working conditions, weak labour and environmental laws, and discrimination; weak states and labour exploitation are rife in the global south, which is the primary supplier of cheap labour. The anti-sweatshop movement in the 1980s and 1990s drew worldwide attention to the plight of workers and the use of child labour in sweatshops supplying shoes and apparel to global brands and retailers. A series of compliance measures emerged from the anti-sweatshop movement such as codes of conduct, multi-stakeholder initiatives (MSIs), CSR programmes, and social and environmental auditing. Given the lack of a uniform global code and the diversity of national regulations and labour laws, these codes of conduct and voluntary standards were seen as ethical compliance measures that could govern the supply chain more responsibly.

In global production networks, manufacturing is outsourced through a system of contracting and subcontracting to low-wage countries without formal employment relations or contracts between the buyers and individual workers. The system of contracting and subcontracting also incrementally reduces the profit margin at each link of the chain, creating further pressures to lower costs of production, resulting in lower wages and adverse working conditions, while reducing accountability and transparency. For instance, through networks of subcontractors in the supply chain, polo shirts made in Bangladesh for $5.67 each are sold in Canada by global retailers for $14, with some workers being paid 12 cents for every shirt they make (Saxena and Labowitz, 2015). Moreover, monitoring worker abuse and violation of worker rights becomes difficult when unauthorized subcontracting is rife, making it difficult to trace and audit every part of the supply chain, as is the case in the Bangladesh garment sector (Siddiqui and Uddin, 2016). The involvement of multiple local and international actors in governing the supply chain – the state, trade unions, NGOs, suppliers, entrepreneurs, global brands and retailers – was an attempt to broaden the primarily buyer-driven dynamic of global supply chains where ‘soft laws’ like compliance standards and codes of conduct could serve as responsible governance mechanisms.

These shifts have also impacted our understanding of state sovereignty as states in the developing regions of the world offering cheap labour become more integrated in complex global supply chains. These states are embedded in multiple transnational networks that serve global markets, and consequently their modes of governing the population have shifted to providing stable and cheap labour. State sovereignty is no longer restricted to the exercise of state power to protect its territorial security but also to refashion the economy and society in order to promote ‘development’ through surveillance, control and regulation (Ong, 2000). In analysing the economies of the so-called ‘Tiger states’ in South East Asia, Ong (2000, 2006) develops the notion of graduated sovereignty to describe practices of governing populations. Graduated sovereignty involves differential treatment of certain populations based on their ability to serve global markets. Fragmentations of class, ethnicity, religion and gender become further intensified as a result. In the case of the Bangladesh garment sector, women constitute the vast majority of workers. The migration of women from rural areas to urban centres and their interpellation into global production networks represent a new class of labour, requiring new forms of surveillance, disciplining and regulation as well as a gradation of citizenship rights. Graduated sovereignty also involves MNCs and international NGOs exerting power and authority through their networks, especially in cases where production is located in special economic zones such as export processing zones. This transition from government to governance is another characteristic of neoliberal globalization, entailing shifting relations between market, state and civil society actors. State regulation is combined with more flexible forms of governance enabling non-state actors to exert power in global production networks.

Ong’s (2000) ethnographic analysis of transnational national production networks in Malaysia and Indonesia showed how state power was reconfigured to meet the demands of global supply chains. The creation of special economic zones along with ‘flexible’ regulatory regimes had differential effects on the rights and entitlements of workers under the flexible labour regime. These states of exception constituted a fundamental condition of neoliberalism in developing countries. However, this does not necessarily imply that postcolonial states have all been weakened; rather, in certain areas of the economy the state has a very strong presence, whereas in others market actors and foreign capital dominate. State repression against dissenting workers, forcible relocation of populations to make way for development projects, and the creation of export processing zones that serve as spaces of exception from national legislation reconfigure the relationship between the state and its citizens, enabling differential treatment of populations in terms of their entitlements and rights.

A similar process can be observed in Bangladesh’s garment industry. The state played a key role in promoting the country’s cheap and flexible labour force as a means to attract foreign capital from corporations that had global supply chains. Despite an active union sector in the country, the power of unions has weakened considerably during the last few decades, which appears to be a global trend. The aim of the state–industry nexus is to align the labour force with global capital requirements of flexibility and low wages, and as a result basic labour rights have been undermined. In addition to state regulations for workers’ rights and conditions, there are also corporate- and civil society-led initiatives and codes of conduct to improve working conditions in garment factories.

CSR and codes of conduct

The early 1990s saw the emergence of the anti-sweatshop movement in the sports footwear industry, where a variety of groups targeted multinationals like Nike and Gap over oppressive labour conditions, alleging repeated violations of core conventions of the International Labour Organization (ILO) (Merck, 2009). The industry response was to develop codes of conduct, monitor working conditions in factories, conduct social and environmental audits, and cooperate with various international auditing standards and schemes. Codes of conduct are CSR initiatives that are also forms of power exercised in global production networks. Structural, relational and discursive power circulates within these networks, establishing particular social relations between actors and creating forms of legitimacy and modes of compliance through specific practices (Clegg, 2014).

As mentioned earlier, following the Rana Plaza disaster, two codes of conduct – the Accord and Alliance – were developed to ensure worker safety in the garment industry. Whereas the former claims to be more legally binding, the latter is framed more as a CSR-driven strategy. Codes of conduct, voluntary standards and social audits do not have a particularly distinguished history: they have been criticized for enabling firms to use deceptive accounting techniques (Frenkel and Scott, 2002), greenwashing (Banerjee, 2011) and, as a practice of self-regulation without sanctions, being subject to opportunism (Lantos, 1999). In some cases, there were no discernible differences in practice between companies that signed up for voluntary standards and those that did not (King and Lenox, 2000).

Studies examining the impact of codes of conduct on working conditions have found that, although there were some improvements in health and safety standards and working conditions, there were no improvements in worker empowerment through better collective bargaining procedures or enhanced freedom of association (Merck, 2009). Critics argue that codes of conduct and standards provide legitimacy to powerful corporate and market actors rather than improve the conditions of workers (Banerjee, 2010). A recent study found that audits of codes of conduct are routinely falsified and, although there have been some improvements in working conditions, there was little evidence that overall levels of compliance had changed (Labowitz and Baumann-Pauly, 2014).

For instance, the Ethical Trading Initiative (ETI), a multi-stakeholder initiative involving companies, trade unions and NGOs, has been focusing on textile workers’ rights in the Indian state of Tamil Nadu since 2012. Despite working with global brands and local manufacturers for five years (and expenditure of nearly £500,000), the ETI’s efforts to promote workers’ rights do not appear to have resulted in any benefits for workers, and neither did they result in greater transparency in the supply chain, according to a 2016 report. The report found that reputation and image management were primary drivers for companies to join ETI, and any improvements in working conditions in factories were carried out as long as they did not affect profit margins (Connor et al., 2016). CSR compliance regimes and codes of conduct operate from asymmetrical power relations that transform social relationships in global production networks between buyers in the global north and actors in the supply chain in developing countries, creating new hierarchies that classify groups and societies based on their ability to comply (De Neve, 2009).

Both state and market failure contributed to the Rana Plaza disaster. The government came under criticism for failing to monitor illegal expansion of the building and failing to close down the building despite local media broadcasting images of cracks in the building and a government inspector’s order to evacuate the building the day before the collapse (Siddiqui and Uddin, 2016). Workers who expressed fears were ‘persuaded and cajoled’ to return to work or threatened with dismissal by factory managers (Human Rights Watch, 2015). None of the five factories operating in the building had a trade union. Days before the accident, two garment factories in the building passed a ‘social audit’ on working conditions based on CSR standards (Donaghey and Reinecke, 2017). Whereas CSR-based programmes such as the Alliance can be seen as providing more legitimacy to market actors and allaying reputational risks, the Accord was designed to represent both worker and business interests. Local employers of the workers – the factory owners – were excluded from the Accord because of their perceived ‘obstructionism’. Instead, market power of global brands was used to overcome the objections of local employers on workers’ rights issues while ostensibly allowing labour unions to play a more prominent role on worker safety issues (Donaghey and Reinecke, 2017).

Thus, it becomes important to study how ideas, practices and legitimacy operate in these compliance regimes in global production networks. Understanding the lived experience of actors that constitute the global supply chain can provide insights into how state, market and civil society regimes operate in the network. So, how were the Accord and Alliance enacted on the ground? What were key actors’ experiences of implementing these governance measures? What was the role of the state in supporting or undermining these measures? In what ways did the mutual constitution of norms intersect with the notion of state sovereignty and the making of graduated sovereignty? To answer these questions, we describe our empirical study in the next section.

Methods

We used a variety of primary and secondary data sources to address our research questions. Secondary sources included newspaper reports, government reports, NGO commissioned reports, reports by human rights activists, industry reports, World Bank and European Commission reports, company reports, reports from research-based advocacy organizations in Bangladesh, research reports from the Centre for Business and Human Rights at New York University, and research reports from the Bangladesh Institute of Labour Studies. In total, we consulted 169 documents and media reports. These documents enabled us to construct a historical narrative of the emergence and subsequent dominance of the garment industry in Bangladesh along with the political contestations arising from worker struggles, union activism, factory owner strategies and practices, international interventions and state policies.

Data collection consisted of a series of in-depth semi-structured interviews conducted by the first author (FA from here on) in three phases between 2013 at the time of signing the Accord and Alliance and December 2016 to January 2017. In total, 18 interviews were conducted, constituting the theoretical sample for the study. Respondents were selected based on their roles in the industry. The FA was born and raised in Dhaka and had prior involvement in the garment sector as a media consultant for the Make Trade Fair Campaign coordinated by Oxfam. Through this experience, she developed an extensive range of contacts in the industry, including activists, labour lawyers, union leaders, factory owners and government officials. Key informants were recruited through her network in the industry. Some respondents suggested other potential informants that were subsequently recruited. For instance, a respondent from the Bangladesh Garment Manufacturers Association mentioned that the Chief Factory Inspector would be a good source of information, and even helped schedule an appointment with the inspector and the FA. Table 1 summarizes the sample.

Table 1.

Sample details.

Affiliation and number of respondents	Details of respondent*
Factory workers (3)	Females (F): Kulsum, Nilu and Runa. Ages range from 15 to 35 years
Union leaders and Accord member (4)	Three females (F) and one male (M)
	Salma (F)
	Zakia (F)
	Rokeya (F)
	Ishan (M)
Labour activists (3)	Deepon (M): member of research-based advocacy organization
	Arif (M): member of research-based advocacy organization
	Mustafiz (M): member of research organization
Labour lawyer (1)	Ali (M)
Investor (1)	Nayeem (M): Executive Vice President of a leading international bank. Interviewed twice during 2014 and 2016
Owners of garment factories (3)	Ruhana (F): Owner of top-tier garment factories. Interviewed twice during 2014 and 2016
	Zafar (M): Owner of mid-tier garment factories
	Sumon (M): Owner of low-tier garment factories
Industry association (1)	Hashim (M): Representative of an industry association
Government officials (2)	Azhar (M): Senior Director
Government officials (2)	Irfan (M): Initially involved with the Sustainability Compact Act
Total = 18

All names are pseudonyms, and details of some job roles have been disguised.

F = female; M = male

Interviews generally lasted between 60 and 90 minutes. They were conducted in Bangla and translated to English by the FA, who is fluent in both languages. In total, 30 hours of recorded material were transcribed and analysed. In addition to interview transcripts, field notes and observations comprising 110 pages in total were kept. The FA spent at least two days a week in the library of Bangladesh Institute of Labour Studies and Integrated Community and Industrial Development in Bangladesh (INCIDIN Bangladesh) during her three field trips. Here she observed the daily informal interactions between activists, trade union leaders, workers and labour lawyers, and participated in conversations where they discussed the Accord and Alliance and the role of the government. The FA maintained detailed field notes throughout the field trips. Discussions during the interviews revolved around current working conditions in factories, whether there were any improvements following the Accord and Alliance, the extent of compliance, participants’ experiences of working in the garment sector, their views on the implementation of the Accord and Alliance, worker struggles demanding an increase in the minimum wage, state repression of dissent, challenges of implementing the Accord, and consequences arising from the shutting down of garment factories. Our overarching question was: how do key actors in the sector – workers, union leaders, owners, activists and government officials – describe the current situation in garment factories? We also sought their views about the Accord and Alliance and whether they felt such governance mechanisms would improve working conditions in the sector, as well as the extent of their participation in the Accord and Alliance process.

Data analysis involved close iterative readings of interview transcripts, field notes, documents and research reports. We initially identified key words, phrases and topics that were frequently mentioned, including workers’ rights, health and safety, subcontracting, fast lead times, pricing and contribution margins, crisis, multiple compliance codes, and state policies. We then grouped these codes into thematic categories by looking for relationships between them, and derived a story line associated with each theme. Our analysis involved looking for links and contradictions between our derived themes as well as insights from the literature on neoliberal state policies, graduated sovereignty through differential treatment of populations, and CSR-based codes of conduct. For example, we found that, because state labour and regulatory policies had to be aligned with the policy prescriptions of the World Bank and the International Monetary Fund to open up the garment manufacturing sector, the regulatory regime in the industry tended to favour factory owners and investors rather than garment workers. This finding was corroborated by comments from labour leaders, who pointed out that there were no wage increases for more than 10 years between 1995 and 2006, whereas factory owners received favourable tax breaks. Thus, we interpreted that the state–market nexus represented a particular arrangement for the neoliberal developmental state that legitimized the state’s differential treatment of its citizens – in this case between workers and factory owners. Workers expressed their collective agency by joining unions, in some cases, which represented their interests against factory owners and the state. In addition, the Accord and Alliance were additional private governance mechanisms to ensure compliance with safety codes. Thus, three themes emerged from our analysis, which we call regimes of the neoliberal developmental state, regimes of representation and regimes of compliance. We elaborate on these themes in the next section.

Findings

Documents, media reports analysis as well as our interviews enabled us to construct a historical timeline of key events in the Bangladesh garment industry as summarized in Table 2, and we elaborate on the themes that emerged from our analysis below.

Table 2.

Timeline of key events in the Bangladesh garment industry.

Year	Events
1978	The RMG sector emerged in Bangladesh with nine export-oriented garment factories. Bangladesh as a potentially key actor in the global supply chain following the quota system under the Multi Fibre Agreement (MFA) with tariff-free access to western markets. National industrial policies broadened the scope for private sector participation in the industrialization process through flexible foreign direct investment schemes and establishment of export processing zones. State-owned textile mills dismantled
1985	First minimum wage law enacted. Minimum wage set at $8 per month
1993	Enactment of Child Labour Deterrence Act
1993	By 1997 almost 50,000 children had been dismissed from the workforce
1994	Revision of wage scale. Minimum wage set at $11 per month
1995	Phasing out of the Multi Fibre Agreement by 2005 by the World Trade Organization
2003	Make Trade Fair Campaign by the local NGOs coordinated by OXFAM GB to secure the livelihood of the workers
2003	BGMEA called for massive government support to make the sector globally competitive following the phasing out of the MFA in the form of investment, low interest rates and cash support
April 2005	Collapse of the nine-storey Spectrum building that housed Spectrum factory in Savar, Dhaka, causing death toll of 70 garment workers
2006	Enactment of first Labour Law of Bangladesh. Key provisions include formal employment contracts, one month notification for retrenchment or discharge, all overtime payments should be double the hourly rate, 8 hour working day, and safe working environment. However, in practice most of these provisions were never enforced
2006	Third revision of the wage scale after 12 years. Minimum wage fixed at $20 per month
2010	Between 2006 and 2010 there were fire accidents in 213 factories causing deaths of 414 workers
	Enactment of the National Building Code
	Fourth revision of the wage scale. Minimum wage fixed at $37 per month
24 November 2012	Fire accident at Tazreen factory, causing deaths of 112 workers
24 March 2013	National Tripartite Plan of Action on Fire Safety and Structural Integrity in the garment Sector of Bangladesh proposed by the ILO
24 April 2013	The collapse of the Rana Plaza Tower in Savar, Dhaka
13 May 2013	Enactment of the Fire and Building Safety Accord with 190 global brands as signatories
	The Accord is endorsed by the Government of Bangladesh, the UN Secretary General, ILO, US Senators and Representatives, European Parliament, OECD National Centre and IndustriALL Global Union and its affiliated Trade Union Federations of Bangladesh
	Enactment of The Alliance for Bangladesh Worker SafetySigned by 28 global retailers in the USA
25 July 2013–2017	Enactment of NTPA under NIP on Fire Safety and Structural Integrity to improve compliance in alignment with the international core labour standards and national labour lawEndorsed by BGMEA, BKMEA, Ministry of Commerce and Ministry of Labour and Employment Affairs
September 2013	Fifth revised wage scale announced and the minimum wage set at $68 per month, which is still lower than a living wage. Garment factory workers’ wages in Bangladesh are still the lowest in the world

BGMEA: Bangladesh Garment Manufacturers and Exporters Association; BKMEA: Bangladesh Knitwear Manufacturers and Exporters Association; ILO: International Labor Organization; NTPA: National Tripartite Plan of Action; NIP: National Integrity Plan; OECD: Organization for Economic Co-operation and Development

Regimes of the neoliberal developmental state

The RMG sector in Bangladesh emerged in 1978 with nine export-oriented garment factories (Labowitz and Baumann-Pauly, 2014). The then government, the first of the country’s military regime, introduced large-scale political and economic reforms and launched major infrastructure projects. A key element in the country’s industrial policy was private sector development and export-oriented growth, and as a result the RMG sector grew rapidly over the next few decades (Muhammad, 2006). Export processing zones (where unions were banned) were also set up during this period to attract foreign investment. The RMG industry experienced exponential growth over the next decades: according to figures from the industry’s primary trade association, the Bangladesh Garment Manufacturers and Exporters Association (BGMEA), there were 384 garment factories employing 120,000 workers in 1984. In 2013, the year of the Rana Plaza collapse, there were approximately 5600 registered garment factories in the country (some estimates mention more than 7000 factories including unregistered indirect sourcing factories) employing nearly 4 million workers, with women comprising nearly 80% of the workforce (Labowitz and Baumann-Pauly, 2014). Almost 70% of the garment factories are located in the capital city Dhaka (Muhammad, 2011).

The industry is Bangladesh’s largest, with annual revenues of $26b in 2015, accounting for 80% of the country’s export earnings and 23% of GDP. Owners of garment factories also became a powerful political ruling class in Bangladesh: at the time of the Rana Plaza accident, 10% of the country’s members of parliament were also owners of garment factories. The industry–state nexus governs garment manufacturing in Bangladesh: a 2013 report of Transparency International Bangladesh found there were strong links between the leadership of the peak industry association BGMEA and the ruling party, which enabled the garment sector to influence legislative processes of the country by creating favourable tax regimes and undermining or blocking health and safety regulations.

While the garment industry is a major contributor to the country’s economic growth and societal development, much of this has come at a cost to workers, especially to their health and safety. It is estimated that more than 3000 garment workers have died in factory accidents in Bangladesh from 1990 to 2013 (Centre for Policy Dialogue, 2013). Although Bangladesh ratified the core conventions of the ILO in 1972 on issues such as collective bargaining, freedom of association and the right to organize, there are still significant shortcomings in the country’s labour laws, particularly in health and safety issues, worker rights in export processing zones and in subcontracting factories. There have been ongoing protests by workers since 2013 over low wages in the sector, where despite five wage revisions since 1985 wages of garment factors in Bangladesh continue to be the lowest in the world. According to labour leaders, the current minimum wage of $68 per month (after a wage revision in 2013) is still not sufficient for workers to make ends meet, whereas factory owners claim that the wage increase is too high. Reflecting on conditions under which they had to work from 1978 to 2013, one respondent, a female trade union leader who was retrenched from her job as a garment factory worker in the 1990s, commented:

From 1995 to 2006 the wage had been 960 taka [US$12]. We had to work 10 years at the same wage. Then it was revised in 2006, because the workers took to the streets and started a movement. The buyers and the donors [international aid agencies] had been aware of all these. They know everything. (Salma)¹

Indirect sourcing via subcontracting through purchasing agents makes regulation of the supply chain extremely difficult. Subcontracting is a key strategy that enables suppliers to maintain low prices, keep wages low, make workers work longer hours (often without overtime pay), and avoid regulatory scrutiny in factories that have dangerous working conditions. Most indirect suppliers are not registered with the government or the two national trade associations, and several buyers turn a blind eye to what is an essentially illegal practice. It is estimated that there are at least 2000 unregistered subcontractors in Bangladesh operating on razor-thin margins, paying below minimum wage, with virtually no compliance with codes of conduct or even local and national regulations (Labowitz and Baumann-Pauly, 2014).

Regimes of representation

Historically in the garment manufacturing sector factory, owners have generally been opposed to unions, often using national laws that require the factory owners’ permission to establish a trade union in their enterprise, in order to block the formation of unions in their factories (Siddiqui and Uddin, 2016). Repression of activists and union organizers through violence and harassment is a common practice in the sector. Labour reforms have done little to guarantee collective bargaining rights or workers’ rights to be associated with a union. Ali, a labour lawyer, commented that in a ‘flexible’ labour regime it was much easier to terminate workers: ‘The owners now can terminate workers by giving the worker 7 days’ notice, and will not be liable to pay compensation’. Another respondent of a research-based advocacy organization commented that there was lack of accountability even after the Rana Plaza tragedy: ‘Only Rana and the owner of Spectrum factory have been sent to prison. Owners are always above the law’. Labour leaders, particularly women leaders, also complained of violence and harassment by police and that workers need protection against Hamla and Mamla (attack and detention). Salma, a trade union leader, commented:

Every owner has their own Mastans [muscle men]; just like in the past the Zamindar [feudal lords] had their own Lathial Bahini [literally ‘armies with sticks’]. In Ashulia [an industrial suburb near Dhaka] the owners of the slums are now controlling the workers as the paid musclemen of the garment factories owners.

The number of unions seeking registration has increased since the Rana Plaza disaster. As of 2015 there were 464 trade unions in Bangladesh of which 323 were registered since the accident (Labowitz and Baumann-Pauly, 2015). However, the country’s labour laws make it difficult to register unions because there is a requirement that one-third of a factory’s workforce has to be registered with a union if it is to receive government recognition. In large factories employing thousands of workers, signing up a third of them can prove difficult, and consequently most of the new union registration is occurring in small factories, where the reach of the Accord and Alliance is extremely limited. The formation of new unions does little to alleviate the historical animosity between factory owners and workers given the state’s long history of violence against union leaders. Although it is considered illegal to block trade union formation, the trade association body BGMEA continues to oppose them, according to several union leaders (Siddiqui and Uddin, 2016). One of the reasons given by factory owners is that trade unions do not really represent the real interests of workers, and are driven by ‘political’ interests. The irony of accusing union leaders of ‘politicizing’ the industry was not lost on a labour advocate, who commented:

The factory owners who occupy 30 seats in the current Parliament are now talking about the ‘politicization’ of industry because of unions? This is ridiculous. They have the political support, they can influence policy formulation process because of their political affiliation, then what is the problem if the workers are united by a union? (Salma)

Silence about the rights of workers in the garment industry to be associated with a union even after the Rana Plaza accident has caused discord among union leaders and labour activists, who had been fighting for unionization since the 1990s. Trade unionism in Bangladesh cannot be delinked from ‘political’ struggles, and unions played a key role during the liberation war of Bangladesh in 1971 and subsequently during the two consecutive army regimes from 1975 to 1990. Several trade unionists joined political parties as a strategy to incorporate workers’ rights into national policies. All four of our respondents from trade unions are members of political parties. According to Ishan, a member of the Steering Committee of the Accord and a union leader, workers in industries other than the garment industry are more exploited: ‘Because the garment industry has global dimensions, so the issue of unionization of the RMG factories workers has drawn more attention comparing to other industries of the country’. However, the women union leaders disagree: ‘These workers bring in dollars as foreign currency into the country, so they should be treated as global workers’ (Salma). The women labour leaders were also critical of the Accord, both in its mechanisms and process of formation, pointing to the exclusion of women during the process:

There are at least 72 trade union federations registered and unregistered. But the Accord and the Alliance never asked our views. Only trade union federations – those that are affiliated with the IndustriALL [a global union federation] – have been involved. (Zakia)

Women workers or labour leaders are not represented either in the steering committee or the advisory committee of the Accord, which is a strange exclusion, to say the least, given that women comprise 80% of the workforce in the sector.

The business model of providing cheap clothing for western consumers puts enormous pressure on suppliers. Demand and fashion styles fluctuate unpredictably at times, and shipment schedules are notoriously tight, which require subcontracting and overtime work (Ali et al., 2002; Kamat, 2016). These factors contribute to accidents in factories. As a respondent from a research-based advocacy organization explained:

As the lead time is shorter the producers have to face two kinds of pressure. They have to force the workers to work on an overtime basis. Similarly, there is a pressure on the capacity of the infrastructure. It facilitated an accident like the collapse of Rana Plaza. (Deepon)

Another respondent, a trade union leader, commented:

When there had been an announcement on the media the previous day that the building [Rana Plaza Tower] might collapse; then why were the workers allowed to work? RMG factories take such an aggressive attitude in order to meet the shipment schedule. The buyers are only concerned about their contribution margin. (Salma)

Workers have no say in assigning targets, although the Accord and Alliance are both committed to worker participation at the factory level. One garment worker commented:

Working in the factory is like living in Dozakh [hell]. There are machines on every floor and there we punch our ID card when we enter the factory. But our problem lies in meeting the targets set by the managers … we are assigned with targets like 1400–1500 pieces and such tasks we have to finish between 8 a.m. and 5 p.m.. So we have to deliver like 150 pieces per hour. If we cannot finish by that time we are not allowed to leave. While working hours after 5 p.m. should be counted as overtime work. But excess working hours to fulfil this assigned target [1400–1500 collars or sleeves] are not considered as overtime work. In fact, management does zulum [cruelty] on us. (Kulsum)

In Kulsum’s factory everyone works 12–14 hour shifts. There are no holiday or leave entitlements except during Eid, one of the most important religious festivals for Muslims. Runa, a 15-year-old factory worker in Dhaka that is not under the jurisdiction of the Accord and Alliance, stated:

In our factory, A shift starts at 8 a.m. and B shift at 8 p.m.. If we come back home at 8 a.m. on Thursday, then on Friday we leave home for work at 7 p.m. [Friday–Saturday constitutes the weekend in Bangladesh so the workers do not get the weekend off]. We are not allowed to talk or joke with our colleagues. If we are suffering from headache we can ask for a break. But if there is pressure on work, to meet targets if we ask for sick leave because we are not feeling well, the management will not approve our request. We have never been told what we have to do if there is a fire in the factory. (Runa)

Other accounts of working conditions mirror the findings from our study. In one case workers had to complete 120–150 pieces of clothing an hour for 14 hours a day until the order was complete. Failure to meet targets resulted in verbal and physical abuse, loss of pay and even sexual harassment (Kamat, 2016). The primary focus of the Accord and Alliance compliance is to prevent accidents and fires in factories and raise health and safety standards. It is doubtful that current practices of long working hours, overtime payments, child labour – all of which are still common practices in Bangladesh’s garment factories, particularly in the unregistered subcontractor factories – can be addressed by these compliance measures.

Although women workers constitute the vast majority of the workforce, they are employed at the lowest pay grades of operators and helpers and sometimes as supervisors, which according to a trade union leader is the highest position women in the factory can obtain: ‘In my entire working life I came across only two female production managers’ (Salma). Following the Rana Plaza collapse, the women trade union leaders were hopeful that discriminatory practices against women would be addressed as the government called on all union leaders to manage the crisis, and for the first time the state acknowledged the role of women union leaders. Not surprisingly, all three women union leaders were convinced that an active trade union at the factory level was key to a sustainable and safe garment sector in Bangladesh. They pointed to the increase in the number of unions at the factory level, which they claim is the outcome of a 25-year struggle to claim workers’ rights. Salma commented:

20 years ago we had to do 25 night shifts in every month. We worked for 10 years under the same wage. During workshops in INCIDIN [a research-based advocacy organization] Masud bhai, Dipu bhai – they used to tell us that we would receive employment contract, our wage would be systematically revised, our right to be unionized, and having a safe working place. Their voice echoed to us as fairy tales. But now we can see these changes … I think now we can solve our problems. (Salma)

The women union leaders also played a role in organizing a demonstration of over 50,000 workers in September 2013 demanding a proper living wage. The demonstration was organized by the trade union federations of the ruling political parties – the Grand Alliance. According to Salma and Zakia, the mass protest served two purposes: ‘First, we gave a message to the owners that the revised wage scale must be followed. And second, the demonstration had a political angle given that the 10th parliamentary elections were due to be held in January 2014’. Unions also attempted to promote the collective agency of factory workers by raising awareness of health and safety issues in the factory and mistreatment of workers, in addition to demands for fair wages.

Although unions can play a key role in influencing production practices and working conditions, the extent of representation is unclear. Under the revised labour law, ‘participation committees’ are required to be formed at the factory level to monitor working conditions. Both Salma and Zakia were concerned that the Accord and Alliance changed practices of unionizing workers. They feel that worker participation in raising awareness of working conditions and securing their future is not a priority in these collective agreements. Prior to the Accord and Alliance, all trade union federations and leaders shared equal responsibilities for giving workers a voice and in resolving pay disputes. Subsequently, only federations affiliated with the Accord and Alliance were entitled to organize workers of those factories under inspection. In the name of promoting inclusion of workers, voices of other unions not affiliated with the Accord or Alliance were excluded.

Union leaders preferred a more direct role for unions. One of the union leaders mentioned it was possible to create an integrated safety committee at the national level consisting of 72 registered and unregistered trade unions. Each trade union would be responsible for monitoring safety and workers’ issues for all of the garment factories located in its geographic zone. She commented: ‘We do not need the Accord and the Alliance. I think we are organized under the trade union, and our code of conduct is our labour law and industrial law – these are enough to establish a vibrant industrial culture’ (Zakia). The union leaders considered it a major achievement of trade union activism that the government revised the labour laws in 2013 after the Rana Plaza accident, and this encouraged them to envision greater union involvement in managing the garment industry. However, four years later when they were interviewed again, they appeared less optimistic. Although the processes of forming unions may have become easier, applications for registering new unions were routinely rejected by the government. According to one union leader: ‘The officials of the Labour Ministry and the owners are in collusion. They see the names [of potential union members] and then verify with the owners and then they provide approval only if the owners approve of the proposed members’ (Rokeya). Such a regime of representation, where essentially factory owners decide on which members serve on the participatory committee, tends to serve the interests of the factory owners rather than the workers.

Ishan, another union leader and member of the Steering Committee of the Accord, while acknowledging the change in law to form unions at the factory level, is sceptical whether these unions or participatory committees could bring about any effective changes in labour relations. He commented: ‘These unions are not organized and are not functioning as collective bargaining agents. Workers do not understand why union is important’. Ali, a labour lawyer, commented that the majority of garment factory workers are first-generation workers, mainly women who migrated from rural areas. According to him, women workers start getting involved with trade unions generally because of pay disputes, and he felt this would eventually help the workers understand the importance of unions on the factory floor.

Nilu, a garment worker, when asked about the role of unions, commented:

Union? What is this? Leader? There is no leader among us who can speak for us. We work and then sometimes we find there is no payment. When suffering becomes intense the men protest. Sometimes women workers also raise their voice. But I find in garment factories the number of employed male workers has been decreased over the years. Because they argue and in some cases they start fighting. We need jobs, and we do not protest: We need work. If we protest there is no work, so we do not protest. If anything happens or for any mistakes, the managers take away our identity card. On the other hand if we cannot swipe on the following day there will be no record of my Hajira [attendance]. And then we can be terminated and management will use the excuse of disobedience. Simply working in a garment factory is Dozakh [hell]. (Nilu)

Whereas the factory owners and government officials felt there were changes at the factory level after the Accord and Alliance and state labour law reform, the workers and their union leaders were more sceptical. One labour leader commented:

Yes, the workers have been provided with uniforms and you can notice such changes instantly, like multiple fire exits and number of toilets But 60% factories still do not provide employment contract. Till now the major problem is excessive work pressure. That has not been addressed. Even if they do night shifts it is not disclosed. Workers ID card is punched and recorded their presence till 7 p.m.. But their work after 7 p.m. is recorded in papers and it won’t be disclosed. This also happens in the Accord listed factories.

One of the provisions of the Alliance was to create a toll-free hotline for workers, who could register their concerns about safety issues. These complaints were investigated by factory managers and Alliance inspectors. However, this initiative serves to individualize problems rather than allow grievances to be voiced collectively. In addition, it was unclear how different concerns raised by workers would be prioritized by the inspectors. Moreover, an overwhelming majority of the users of these hotlines were men, which raised concerns about this particular regime of representation, given the predominantly female composition of the workforce (Donaghey and Reinecke, 2017).

Regimes of compliance

The Accord and Alliance are compliance measures that were developed to ensure health and safety in the sector and provide transparency, measurable results and verifiable changes in the garment sector. The Accord’s key decision-making body is a steering committee comprised of three representatives selected by trade union members and three representatives selected by retailer participants. The ILO selects a neutral chair. The Accord also has an Advisory Board with representations from the Ministry of Commerce and the industry association BGMEA. Their task is to create a list of all of the suppliers of the companies, prepare semi-annual or quarterly inspection reports, and regularly publish and update data on factory inspections and training programmes. There are no representatives from factory owners and garment manufacturers in the Advisory Board. The Alliance’s governing body consists of four brand representatives, four ‘outside experts’ and an elected chair. It also has a board of advisors comprising 12 multi-stakeholder industry experts. Funding to cover inspections and training is provided through annual contributions by buyers, but it is the responsibility of factory owners to pay for structural repairs or remediation where needed. More than 200 international brands along with global and local unions are signatories to the Accord, whereas the Alliance comprises 27 North American brands. Both initiatives have received significant international public attention and have committed to spend up to US$100m over five years to improve factory safety. Both have a five-year commitment, which will end in 2018, with timelines and accountability for conducting safety inspections, providing fire and building safety training and worker empowerment programmes.

Initially, the Accord and Alliance were received favourably by key actors in the garment industry. The Accord categorizes garment factories as top tier (A), mid-tier (B) and low tier (C) based on the degree of compliance of the factories with the building and safety codes, their annual turnover and the number of workers employed in the factories. Zafar, the owner of a B category factory, stated:

Usually, buyers impose their own compliance mechanisms, so it varies from one buyer to another buyer, and it becomes difficult for us to follow. The compliance consideration of H&M is different from Zara. So, we require a common compliance framework.

Ruhana, the Managing Director and the owner of an A-category factory, suggested that these agreements symbolize the commitment of global brands to the industry: ‘We had been asking for such commitment from them for the last 20 years’. Hashim, an industry representative commented: ‘[T]he way media had been publicizing the image of the industry, there had been assumptions that the industry would not survive without global compliance measures, so we were supportive and positive about the enactment of the agreements’. Another trade union leader who was involved with the Accord since its inception commented:

Media had been publicizing messages like there are outlets that would not sell clothes procured from Bangladesh; and so our negotiation with the buyers centred on this issue. We told them, look you have been doing business with us for 25 years, so you have responsibility … so many people are employed in the industry. (Ishan)

Although the Accord and Alliance represent an important step in the direction of improving worker safety in the region, their actual impact is limited, according to several reports (Labowitz and Baumann-Pauly, 2014). Both the Accord and Alliance cover only 27% of Bangladesh’s garment factories and about 45% of its workers. Most indirect suppliers and subcontractors fall outside the scope of the Accord and Alliance’s inspection zones – about 2000 factories are expected to be inspected under both schemes, which excludes a majority of factories, especially the smaller ones where conditions are much worse and where workers are most vulnerable. Factory owners routinely showcase ‘good’ factories with higher safety standards to foreign buyers, but most of the production is subcontracted to smaller factories and indirect subcontractors who avoid inspection. The compliance regime is essentially a short-term inspection programme for larger factories, leaving the source of the problem unaddressed.

Moreover, the extent of remediation in factories that have been designated unsafe is extremely low. Factory owners are unwilling or unable to bear the high costs of remediation ($250,000 to $350,000 per factory), which according to the terms of the Accord and Alliance is the owner’s responsibility. The Accord and Alliance are compliance regimes that inspect and categorize factories but do not fix the identified problems. One industry representative described the problem of costly compliance:

Okay as an example they [the Accord] asked us to install a fire door. We installed. Later they said that the door specification had to be certified by their assigned third party. Then same thing happened with hydrant and water pump, and other items. Certification means invariably the cost of that product would escalate. This has happened to some extent with the Alliance too. Yes we have learnt how to make a factory compliant. But we had to bear the cost of their decisions. We didn’t receive a single farthing. (Hashim)

Factory owners, union leaders and government officials also raised concerns about the ethics and governance of the Accord and Alliance. They pointed to their lack of inclusion in the governance of the Accord, lack of clarity about funding for remedial measures in factories that were deemed unsafe, and the real possibility of loss of buyers owing to factories being non-compliant. Zafar, the owner of a mid-tier factory, commented: ‘How have they got power to shut down factories? What is their authority?’

One factory owner claimed he invested more than $125,000 to make his factory compliant by borrowing from his bank: ‘If I made $1 profit out of each product, now this profit margin is almost 70–60 cents. But the buyers are offering the same procuring price. Then who bears the cost of compliance?’ Another respondent, a banker, commented: ‘I think they [the brands] are extracting the cost of this compliance programme from here [Bangladesh]. Perhaps they treat the compliance initiatives as a part of their CSR programme, but you know … in the end this is business’. Remediation measures for noncompliant factories can increase the production cost by 13% (The Daily Samakal, 2014).

The responsibility of bearing costs of compliance is a key issue. On the one hand, global brands and retailers want cheap clothes from developing countries. Prices are low because wages are low and working conditions do not meet international standards. Most factory owners are unable to bear the costs of remediation. Unless suppliers are able to charge higher prices (which means that the buyer will seek other cheaper sources) or are provided sources of finance (with which the Accord and Alliance claim they can ‘assist’), no meaningful remediation can take place. Factory owners we interviewed claimed that the Accord had failed in upholding its financial commitment of providing access to low-cost loans to support remediation measures. Sumon, a category C garment factory owner, remarked:

Funding is the most critical issue for us. The entire business is based on credit and suppliers can get funds from the bank if we receive an order. But if there is a rejection of an order, the suppliers have to pay a penalty charge of up to 40%. We do not have sovereignty. We have sold that out. From 1990 to 2003 we had the capacity to negotiate. However, now we have lost that power. We have lost our negotiation capacity. Now Navana, Otobi, and Square – such big Bangladeshi corporate houses – are interested in investing in this Sector. They have the ability to meet the compliance codes and can go for environmentally sustainable production. (Sumon)

One of the provisions of compliance is that suppliers must own their factories and not operate out of rented premises, which has significant implications for existing owners. Zafar, who was initially supportive of the common compliance measures, commented: ‘My factories are located in a rented building, and I already invested in the plant. Again, the bank charges 17% interest rate; and now the wage scale has been revised. I am planning to sell off one of my factories’. Union leaders and labour activists were less sympathetic to the owners’ plight, claiming the latter always raised cost issues when it came to wages. One respondent from a research organization stated, ‘The owners negotiate with the government – and get benefits out of our paid tax. If this industry is the lifeline of the economy of the state then … they [owners] should be prepared for facing such challenges’ (Arif).

Respondents also questioned the governance approach of the Accord. An industry representative commented:

It [the Accord] does not reflect features of a tripartite system. We were asked to be in the Advisory Committee. As a member of the Executive Committee of the BGMEA we are informed about their activity through routine reports. In the Advisory Committee, again our task is to simply listen what they have done. So we have stopped attending that meeting. But the Steering Committee is a policy making committee, where we wanted to be included. But they told us ‘if you want to be included you cannot express opinions that are different from the committee’. That is why we are not going to accept any extension of both the Accord and the Alliance beyond 2018. The Accord is a brand led multi-stakeholder forum, where ILO is merely involved as an umbrella institution. Its major concern is health and safety issues. And we have already communicated to the government how we would like to see the approach of managing this sector beyond 2018. There should be a single body and that body would be responsible to set the code of conduct of the Sector. That platform should include the representatives of buyers, suppliers, BGMEA and BKMEA’s representation, workers, government and ILO. (Hashim)

However, a trade union leader dismissed BGMEA’s views as representing the vested interests of factory owners: ‘We need the Accord Body … Do you think the Factory Inspectors of Bangladesh can independently inspect any factories?’. The description by the BGMEA representative of the Accord as a ‘brand led multi-stakeholder forum’ seems to imply little more than a symbolic commitment by the ILO as an ‘umbrella institution’ within the Accord.

Demonstrations and protests by garment workers broke out in December 2016 sparked by reports that a woman working in a garment factory was physically abused for requesting maternity leave. The workers were also demanding a further revision of the wage scale because of spiralling living costs. These protests were violently repressed by the state, and several union leaders went into hiding to prevent being arrested. One union leader commented:

The owners’ have closed down 55 factories. The police attacked workers’ slums and have been arresting us … They have threatened the workers that they would visit the workers’ homes in their villages and ask their parents to force them to surrender. See how they terrorize us … If the Accord and the Alliance have brought magical changes … then why they [workers] launched a demonstration when a woman worker had been physically abused for asking maternity leave? The Accord and the Alliance have served the interests of only a handful of people involved with the industry. (Rokea)

Another union leader and a member of the Accord commented:

Only through forming union and through active union workers’ awareness can be raised and transparency can be ensured. This [raising awareness of rights] was not an objective of the Accord. The Accord has its limitations because safety and health are its core issues. (Ishan)

By January 2017, the government was successful in suppressing the protests and demonstrations, and the closed factories began operating again while workers continued to live with increasing levels of insecurity and fear of termination. In particular, all the workers who had organized protests over working conditions and pay or tried to form or join a union lost their jobs when the factories reopened and were unable to find other jobs in the sector as the state and factory owners had warned other factory owners about their union activity.

Discussion and concluding remarks

Our analysis of the garment industry in Bangladesh following the signing of the Accord and Alliance has attempted to provide perspectives from the target populations of the compliance measures: factory owners, trade associations, workers and union leaders. In comparing the functions of the Accord and Alliance, Donaghey and Reinecke (2017: 2) concluded that the Alliance was a CSR-based approach that could ‘foster effective problem-solving in the short run’ whereas the Accord was an ‘industrial democracy’ model that was ‘necessary to build governance capacities involving workers in the long run’. Our findings indicate that worker involvement in the Accord and Alliance is limited and tokenistic and does little to address exploitative working conditions. We believe it is both premature and dangerous to claim that the Accord is a victory for labour rights in Bangladesh. It may be a victory for elite, well-funded federations of unions and NGOs in the global north who are key stakeholders in the Accord. It may also be a victory for global brands and retailers because the Accord provides an additional sheen of legitimacy to their glossy sustainability and CSR reports. The industry–state nexus in Bangladesh can now point to the Accord to demonstrate that the industry and the state takes worker safety issues seriously by showcasing larger factories that have implemented new safety procedures, while the dangerous and exploitative working conditions in the vast majority of subcontractors and smaller factories remain unchanged. Providing information about workers’ rights to workers who are unable to exercise those rights because of oppressive labour conditions does not constitute empowerment of workers.

Our findings cast serious doubts about the effectiveness of the Accord in empowering workers, who are the most vulnerable actors in the supply chain in global production networks. In the aftermath of the Rana Plaza tragedy there were external pressures from trade unions, NGOs and global brands to improve worker safety as well as internal demands from workers and unions to raise wages and provide better working conditions. However, despite some level of union representation in the Accord, both the Accord and Alliance remain brand-led initiatives. The Accord does not specify mechanisms for worker participation at the factory level, and as a result factory owners and the state have captured the process and dictated the composition of proposed participation committees. Genuine worker empowerment and welfare comes at a price that has to be paid by buyers. If contribution margins continue to determine what can be spent on worker safety, then the Accord, far from being an initiative of ‘industrial democracy’, is merely a CSR compliance initiative. All evidence post-Rana Plaza suggests there has been no increase in the procurement price for factory owners. In fact, the price of Bangladesh’s major clothing export to the US has dropped 46% since 2000. Nine percent of that drop occurred after the Rana Plaza collapse (Kamat, 2016). Building and fire safety codes following the Accord and Alliance may well reduce factory accidents, but the structural exploitative labour conditions that jeopardize worker safety remain unaddressed.

As the initial tenures of the Accord and Alliance come to an end it is worth reflecting on what they have achieved. There is little evidence that suggests the compliance mechanisms can effectively fill the ‘regulatory vacuum in global governance’ (Scherer and Palazzo, 2011: 899). Multi-stakeholder initiatives like the Accord and Alliance provide legitimacy to powerful actors but do little to address the sourcing, pricing and procurement practices in the garment industry, which are the key reasons for the deplorable condition of workers and that remain outside the mandate of these compliance mechanisms. Factory inspections under the Accord and Alliance do not address the root causes of unsafe working conditions and have done little to change the culture in garment factories. Moreover, global brands escape accountability in cases of violation of workers’ rights, even in cases where factories that manufacture branded garments have ‘passed’ inspection (Saxena and Labowitz, 2015).

Most of the value from the supply chain of garment manufacturing is captured by the big western brands and retailers, with suppliers and workers receiving minimal value (Lund-Thomsen and Lindgreen, 2014). Unless the basic sourcing model is reconfigured and prices reflect fair wages and good working conditions, improvements in the industry will be merely cosmetic. The Accord and Alliance do little to grant workers agency in determining their working conditions, while the state–industry nexus ensures that the labour market remains ‘flexible’ and ‘competitive’ to the needs of the global market, which essentially means the lowest possible wage and minimum investments in factory safety so as to ensure the procurement of orders at the lowest possible price. Codes and standards are not just technocratic solutions, they also ‘generate new social regimes of power and inequality’ (De Neve, 2009: 63) and, as we have seen through regimes of neoliberal development, representation and compliance provide buyers with new forms of legitimacy where ethical compliance becomes a tool for negotiation in addition to price and quality. A measure of success for the Accord and Alliance appears to be the number of unsafe factories that have been shut down. According to government reports, 39 factories have been closed down and 40 factories are ‘partially closed’, affecting nearly 21,000 workers. One newspaper report mentions that 150,000 workers have been affected as a result of closures, including factories that exited the business (Hossain, 2014). The resultant loss of livelihood leads to further disempowering consequences for women workers, many of whom migrated from rural areas and who risk being trafficked to neighbouring countries (Ali et al., 2002).

The Accord and Alliance have reshaped relations between buyers in the global north and suppliers in the global south, but the costs of compliance are borne by actors in the supply chain who do not have the resources to do so. Health and safety outcomes in the garment manufacturing sector in Bangladesh should not be decoupled by their social relations that shape specific labour regimes, but in essence this is what the Accord and Alliance is doing rather than challenging embedded norms and practices of sourcing and subcontracting or providing opportunities for workers to collectively negotiate their rights (Barrientos and Smith, 2007). Our findings show that, whereas workers conceive agency as collective forms of resistance, their agency at the individual level is constrained by the indirect sourcing model and the ultracompetitive piece-rate payment nature of subcontracting (Carswell and De Neve, 2013). The politicization of MNCs is reflected in their ability to exert power over global production networks by creating a private governance framework designed to improve working conditions in factories. The inclusive nature of multi-stakeholder initiatives like the Accord and Alliance through the participation of market, state and civil society actors is assumed to deliver positive outcomes in terms of factory safety and working conditions. However, our findings indicate that key actors involved in these compliance measures operate very much within the mandate prescribed by global brands and retailers. Munir et al. (2017) reported similar findings in their analysis of global production networks in Pakistan’s garment industry, where transnational managerial elites played a key role in transforming governance regimes that served to protect the interests of western brands rather than those of workers.

Development in Bangladesh has followed a neoliberal trajectory since the imposition of structural adjustment policies since the 1980s, and the garment industry is a key driver of development. Reconfiguration of state institutions to serve capital is a condition for a neoliberal developmental state to be embedded into global supply chains. Our findings show that this reconfiguration also involved use of state violence to suppress dissent while undermining labour rights and working conditions. Neoliberalization, as Harvey (2005) points out, also creates the conditions for formations of new classes, as was evident in Bangladesh first in the creation of the military-political class and subsequently the industry-political class, where factory owners became part of the ruling political class in the country. Our findings show how incorporation of local economies into global supply chains resulted in graduated sovereignty in Bangladesh. Neoliberal development policies of the state resulted in differential treatment and regulation of specific populations comprising garment factory workers. A new class of workers – mainly women who migrated from rural areas – emerged as providers of labour, which also transformed social relations of production under neoliberalism. Women who became primary income earners in their household were now governed and disciplined by new regimes of capital and development, further fragmenting citizenship (Ong, 2006). Other classes of people in the sector – owners of large factories, for example – received differential treatment from the state in the form of incentive and tax breaks to create a more investor-friendly climate. However, it is important to realize that, despite significant power imbalances between local and global actors in the supply chain, workers are not passive ‘victims’ of global production networks but actively exercise their agency through collective organizing, protests and demands from the state.

State actors also collaborated with business and non-state actors to create both ‘hard’ and ‘soft’ laws that invariably protected their own interests. In the case of the Accord and Alliance, CSR and stakeholder engagement provided the legitimacy for global brands and retailers. However, the underlying causes of labour exploitation in the sector remain unaddressed by these compliance measures. In the new era of ethical compliance, the global production system remains unchanged except for new standards and codes of conduct that can never be implemented because of a lack of necessary resources and political will of the industry–state nexus. Appropriation of value from the global supply chain still remains in the hands of providers of capital, global brands and buyers, and the state, despite claims of worker empowerment. So here, then, is the legitimacy paradox: under the Accord and Alliance, will it still be possible for a global retailer to sell a branded T-shirt for $10 (perhaps with a buy-one-get-one-free promotional offer) when the women who manufactured the shirt are paid $2 a day and continue to work in unsafe and oppressive conditions?

Footnotes

Acknowledgements

We are deeply indebted to the workers and the leaders of the trade union federations involved with the ready-made garment sector of Bangladesh. We would like also to thank AKM Masud Ali, Rafiqul Islam and Syed Sultan Uddin Ahmmed. Without their support this research would not have been possible. Finally, we would like to thank the two anonymous reviewers of our article.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Notes

Fahreen Alamgir works as a lecturer in the Department of Management, Monash University, Australia. She received her Master’s Degree in Management from the University of Dhaka and her PhD from the School of Management of RMIT University, Melbourne, Australia. Her research examines organizational challenges and mechanisms from the perspectives of social justice, and rights and capability, of the involved and affected community in response to globalization. She is currently researching the cheap clothing industry’s global supply chain, with a particular focus on workforce feminization. She draws from critical political economy, subaltern studies and postcolonial feminist studies. [Email: fahreen.alamgir@monash.edu]

Bobby Banerjee is Professor of Management and Director of the Executive PhD Programme at Cass Business School, City University of London, UK. He received his PhD from the University of Massachusetts at Amherst, USA and has held academic positions at the University of Wollongong, RMIT University, University of South Australia and University of Western Sydney. His research interests include sustainability, climate change, corporate social responsibility, critical management studies and indigenous ecology. He has published widely in international scholarly journals, and his work has appeared in Academy of Management Learning & Education, Business Ethics Quarterly, Human Relations, Journal of Business Research, Journal of Management Studies, Journal of Marketing, Management Learning, Organization and Organization Studies. He is the author of two books: Corporate Social Responsibility: The Good, The Bad and The Ugly and the co-edited volume Organizations, Markets and Imperial Formations: Towards an Anthropology of Globalization. Bobby is a Senior Editor at Organization Studies. [Email: bobby.banerjee@city.ac.uk]

References

Ali

M AKMa

Ali

M AKMu

Sarkar

(2002) The RMG Sector of Bangladesh and its Female Workers: Awaiting a New Global Movement. Dhaka: INCIDIN Bangladesh.

Anwary

(2017) Feminised workforce in transnational production: The Bangladesh readymade garment industry. History and Sociology of South Asia 11(2): 1–18.

Banerjee

(2010) Governing the global corporation: A critical perspective. Business Ethics Quarterly 20(2): 265–274.

Banerjee

(2011) Embedding sustainability across the organization: A critical perspective. Academy of Management Learning & Education 10(4): 719–731.

Barrientos

Smith

(2007) Do workers benefit from ethical trade? Assessing codes of labour practice in global production systems. Third World Quarterly 28(4): 713–729.

Carswell

De Neve

(2013) Labouring for global markets: Conceptualising labour agency in global production networks. Geoforum 44: 62–70.

Centre for Policy Dialogue, CPD (2013) First Independent Monitoring Report: 100 Days of Rana Plaza Tragedy: A Report on Commitments and Delivery. Dhaka: CPD.

Clegg

(2014) Circuits of power/knowledge. Journal of Political Power 7(3): 383–392.

Connor

Delaney

Sarah

(2016) The ethical trading initiative: Negotiated solutions to human rights violations in global supply chains? Non-judicial Redress Mechanisms Report Series No. 18. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2889121 (accessed 25 May 2017).

10.

De Neve

(2009) Power, inequality and corporate social responsibility: The politics of ethical compliance in the South Indian garment industry. Economic and Political Weekly 44: 63–71.

11.

Donaghey

Reinecke

(2017) When industrial democracy meets corporate social responsibility – a comparison of the Bangladesh Accord and Alliance as responses to the Rana Plaza disaster. British Journal of Industrial Relations, https://doi.org/10.1111/bjir.12242.

12.

Frenkel

Scott

(2002) Compliance, collaboration, and codes of labor practice: The Adidas connection. California Management Review 45: 29–49.

13.

Harvey

(2005) A Brief History of Neoliberalism. Oxford: Oxford University Press.

14.

Hossain

(2014) About 400 RMG factories shut from loss of orders, compliance issues. The Daily Independent, 10 September. Sourced from the Archive of Bangladesh Institute of Labour Studies.

15.

Human Rights Watch (2015) “Whoever raises their head suffers the most:” Workers’ rights in Bangladesh garment factories. Available at: https://www.hrw.org/report/2015/04/22/whoever-raises-their-head-suffers-most/workers-rights-bangladeshs-garment (accessed 25 May 2017).

16.

Kamat

(2016) “We are nothing but machines to them”. Available at: http://www.slate.com/articles/business/the_grind/2016/12/bangladesh_s_apparel_factories_still_have_appalling_worker_conditions.html (accessed 25 May 2017).

17.

King

Lenox

(2000) Industry self-regulation without sanction: The chemical industry’s Responsible Care Program. Academy of Management Journal 43: 698–716.

18.

Labowitz

Baumann-Pauly

(2014) Business as usual is not an option: Supply chain and sourcing after Rana Plaza. Center for Business and Human Rights at NYU Leonard N. Stern School of Business. Available at: http://www.stern.nyu.edu/sites/default/files/assets/documents/con_047408.pdf (accessed 25 May 2017).

19.

Labowitz

Baumann-Pauly

(2015) Beyond the tip of the iceberg: Bangladesh’s forgotten apparel workers. Center for Business and Human Rights at NYU Leonard N. Stern School of Business. Available at: http://www.stern.nyu.edu/sites/default/files/assets/documents/con_047408.pdf (accessed 25 May 2017).

20.

Lantos

(1999) Motivating moral corporate behavior. Journal of Consumer Marketing 16: 222–233.

21.

Levy

(2008) Political contestation in global production networks. Academy of Management Review 33(4): 943–963.

22.

Lund-Thomsen

Lindgreen

(2014) Corporate social responsibility in global value chains: Where are we now and where are we going? Journal of Business Ethics 123(1): 11–22.

23.

Merck

(2009) Jumping scale and bridging space in the era of corporate social responsibility: Cross-border labour struggles in the global garment industry. Third World Quarterly 30: 599–615.

24.

Muhammad

(2006) Globalisation and economic transformation in a peripheral economy: The Bangladesh experience. Economic and Political Weekly 41: 1459–1464.

25.

Muhammad

(2011) Wealth and deprivation: Ready-made garments industry in Bangladesh. Economic & Political Weekly 46: 23–27.

26.

Munir

Ayaz

Levy

Willmott

(2017) The role of intermediaries in governance of global production networks: Restructuring work relations in Pakistan’s apparel Industry. Human Relations, https://doi.org/10.1177/0018726717722395.

27.

Ong

(2000) Graduated sovereignty in South-East Asia. Theory Culture and Society 17(4): 55–75.

28.

Ong

(2006) Neoliberalism as Exception: Mutations in Citizenship and Sovereignty. Durham, NC: Duke University Press.

29.

Saxena

Labowitz

(2015) Monitoring working conditions at factories won’t stop future tragedies. The Globe and Mail, 10 August.

30.

Scherer

Palazzo

(2011) The new political role of business in a globalized world: A review of a new perspective on CSR and its implications for the firm, governance and democracy. Journal of Management Studies 48: 899–929.

31.

Siddiqui

Uddin

(2016) Human rights disasters, corporate accountability and the state: Lessons learned from Rana Plaza. Accounting, Auditing & Accountability Journal 29(4): 679–704.

32.

The Daily Samakal (2014) The collective compliance and challenges for the ready-made garment factories. 8 September. Sourced from the Archive of Bangladesh Institute of Labour Studies.

33.

van der Heijden

Zandvliet

(2014) Enforcement of Fundamental Labor Rights. The Network Approach: Closing the Governance Gaps in Low-Wage Manufacturing Industries. Policy Brief 12. Available at: http://www.thehagueinstituteforglobaljustice.org/wp-content/uploads/2015/10/PB12-Enforcement-Labor-Rights.pdf (accessed 25 May 2017).