Labour rights from labour wrongs? transnational compensation and the spatial politics of labour rights after Bangladesh’s Rana Plaza garment factory collapse

The collapse of the Rana Plaza garment factory building in Bangladesh, which resulted in the loss of at least 1,134 lives and injuries to hundreds more, exposed the brutality of a global production system in which labour rights have become privatised, circumscribed, and deterritorialised. Unprecedented for an incident of its kind, affected families received $30 million in “compensation” from global apparel companies through the Rana Plaza Arrangement, a voluntary initiative overseen by the International Labour Organization (ILO). This article situates the Arrangement in debates over labour rights in global supply chains, presenting it as a “hybrid” mechanism that recognises workers’ right to compensation for injury or death but relies on the voluntarism of corporate social responsibility for funding. Exploring the complex role of neoliberal regulation in reproducing geographies of uneven development, I show how a transnational initiative can restrict labour rights even as it attempts to expand such rights.


Introduction
The April 24, 2013 collapse of the Rana Plaza garment factory building in Dhaka resulted in the loss of at least 1,134 lives and injuries to hundreds more. Following several other deadly factory collapses and fires in Bangladesh, the Rana Plaza collapse was an extreme manifestation of the substandard labour conditions and threats to safety that have plagued the global garment industry. The collapse illustrated the dangers of a global production system in which subcontractors compete in a race-to-the-bottom, and multinational companies (MNCs) are not made liable for conditions in the factories producing their goods. With few exceptions, MNCs have evaded all accountability for previous factory fires and collapses (Miller 2012). Yet, within the crucible of worldwide attention in the post-Rana Plaza period, injured workers and the families of those killed received $30 million in compensation for death and injury from global apparel companies through the Rana Plaza Arrangement, a multi-stakeholder agreement overseen by the International Labour Organization (ILO).
The transnational and Bangladeshi NGOs and trade unions that spearheaded the campaign for compensation have described the Rana Plaza Arrangement as a labour rights victory (Clean Clothes Campaign 2015;Kelly 2015;Prentice 2018). Although global apparel companies donated funds voluntarily without any admission of responsibility for the disaster, the amounts paid to injured workers and bereaved families met the international standards of worker compensation for occupational injury or death specified in the ILO's Employment Injury Benefits Convention No. 121 (1964/1980. Compensation was therefore awarded based on claimants' status as rights-bearing workers (or the dependants of workers), and not simply as charity directed towards victims of a shocking and widely-publicised industrial disaster.
Labour rights have long been under assault in the global garment industry (Hale and Wills 2005). Defence of labour rights is historically a state responsibility, and global supply chains stretch across multiple jurisdictions, creating predicaments of regulation and enforcement that supranational institutions, such as the ILO and World Trade Organization, were not designed to adjudicate (Maryanov 2010;Seidman 2012;Standing 2008). Yet, the assault on labour rights is also a basic feature of what global supply chains do: they produce spaces of accumulation where labour is actively devalued (Katz 2001;Selwyn 2012;Tsing 2009). Global production relies upon regulatory unevenness in the social and spatial divisions of labour, where the reproduction of racialised and gendered difference is structurally essential (Werner 2016;Wright 2006). In Bangladesh, the state's role in these processes is evident in policies to attract capital by denying workers' rights to freedom of association and trade union formation, legislating minimum wages that are not living wages, and neglecting to inspect and regulate factories (Ashraf and Prentice 2019;Muhammad 2015;Saxena 2014).
By calculating claims for compensation via ILO standards, the Rana Plaza Arrangement recognised workers' entitlement to rights that had been denied. This reassertion of labour rights in the face of their extreme suppression and neglect made the Arrangement unique as a compensation agreement, both worldwide and in comparison to previous post-disaster compensation deals between MNCs and trade unions in Bangladesh, which I discuss in more detail below (cf Maher 2012, Miller 2012. The Rana Plaza Arrangement is best understood as a hybrid mechanism that recognises workers' limited rights to work injury compensation-a right that would normally be established in law and guaranteed by the state-but relies on the voluntarism of corporate social responsibility (CSR) for funding. 1 As such, it embodies the neoliberal approach to labour governance that dominates the global garment industry today (LeBaron and Lister 2015; Meardi and Marginson 2014). Because the Arrangement has since become a model for awarding compensation after other factory disasters in South Asia (Sumon et al. 2017, Prentice 2018, it raises important questions about how labour rights are being imagined and reconfigured in response to neoliberal regulatory failures, particularly when initiatives claiming to promote labour rights replicate the same neoliberal logics. In this article, I present research with the designers of the Rana Plaza Arrangement-transnational actors from civil society, business, and government-to understand how "labour rights" are constituted in elite spaces with consequences for their deployment in the national context of Bangladesh. Ethel Brooks' (2007) study of transnational labour organising shows how privileged sites of "global contestation" involving activists and MNCs shape labour practices to the exclusion of workers and their everyday experiences. The Rana Plaza Arrangement was mostly formulated at the ILO offices in Geneva and Dhaka, the two main "sites" of my research. I analyse internal documents of the Rana Plaza Arrangement that are available in the public domain, along with press releases, statements, and reports published by the institutions and actors involved in design and implementation. My analysis of these materials is supplemented by 21 semi-structured interviews with key participants in the Rana Plaza Arrangement between October 2016 and November 2019. Interviews lasted between one and three hours and were conducted via Skype, or in person in Geneva, London, Brussels, or Dhaka. Interviewees included members of the tripartite governing body, the Rana Plaza Coordination Committee (RPCC); corporate representatives of brands that contributed to the Rana Plaza Donors Trust Fund; national and international policymakers, including ILO officials; and representatives of trade unions and NGOs based in the global North and Bangladesh, including European labour rights activists who campaigned for apparel brands to donate funds.
In conducting research with these individuals, I presented myself as a labour scholar seeking to understand the story of their participation in the Rana Plaza Arrangement, including their commentary on its successes, failures, and wider implications. Though I sought to capture informants' experiences in their own words, I did not present myself as a purely disinterested observer. I explained that my previous research told a story of globalisation from workers' perspective (Prentice 2015); some informants from activist circles knew me through friends and colleagues working in academia or labour rights NGOs in Bangladesh and Europe. My focus here is on the design of the Arrangement-which was created without input from workers, other than via high-level representatives from Bangladeshi national trade union federations-rather than the experiences of claimants. This article demonstrates that in the variegated policy landscape of the global garment industry, elite actors are embedded within, and can unintentionally reproduce, geographies of regulatory uneven development. My analysis of the Arrangement as a mechanism of neoliberal governance shows the potential for a transnational initiative to restrict labour rights even as it attempts to expand such rights.

Labour Rights in Global Supply Chains
The collapse of the Rana Plaza building on April 24, 2013 was the deadliest disaster in the history of garment manufacturing. The eight-storey building in the Savar industrial district of Dhaka housed on its upper floors five garment factories that produced clothing for global brands such as Walmart, Primark, Benetton, The Children's Place, and Joe Fresh (Clean Clothes Campaign 2013). Cracks appeared on the building the day before the collapse, and Rana Plaza was evacuated and inspected by engineers who recommended that the building remain closed. Banks and retail spaces on the lower floors did not reopen on the 24 th , but garment workers were made to return to work because of production deadlines (Muhammad 2015: 146;Siddiqui and Uddin 2016: 693). When an ordinary loss of electrical power stimulated four diesel generators in the back of the building at 8:45 a.m., the building collapsed, killing at least 1,134 individuals, and trapping and injuring several hundred more (Motlagh 2014: 66). Investigations later determined that the building was constructed with substandard materials and lacked proper inspection and approval (Siddiqui and Uddin 2016: 693).
Activists and labour scholars insist that the despite the enormity of the Rana Plaza disaster, it should not be interpreted as an extraordinary or "exceptional" event (Mezzadri 2017). The collapse is consistent with the ongoing realities of the global garment industry, where multinational brands and retailers demand high levels of production quality and speed at extremely low prices. Particularly with the accelerated pace of "fast fashion," the purchasing practices of MNCs pit garment factories against one another, disincentivising local employers from investing in factory standards (Bulut and Lane 2011;Muhammad 2015). Buyers' fluctuating demands manifest in long working hours, forced and extreme overtime, and the pressure to work even in visibly unsafe conditions. As a result, workers are employed within what Alessandra Mezzadri (2017) has called a "sweatshop regime" of low wages, long hours, and degraded conditions. Recognising that these labour practices are persistent-and not just a temporary stage for garment-exporting countries like Bangladesh-demands investigation into how global production relies upon and reproduces capitalism's uneven development (Werner 2016: 6-10).
To understand why workers entered an unsafe building that day requires attention not only to the global sourcing regime but also to the nexus of MNCs, local factory owners, and the Bangladesh state (Saxena 2020). As Javed Siddiqui and Shahzad Uddin (2016) argue, in Bangladesh the government is so entwined with its garment industry-a sector that provides 80% of its foreign exchange earnings-that they mutually reinforce each other's economic and political power and collude to suppress labour rights. The state's commitment to the export garment industry entrenches poor labour standards because industrial expansion is prioritised over social upgrading (Ahmed et al. 2014;Ashraf and Prentice 2019). The suppression of freedom of association and trade union formation is "justified as a valid 'cost' of maintaining the nation's competitive edge" (Siddiqi 2017: 62). The owner of Rana Plaza, Sohel Rana, was a leader in the youth wing of the ruling political party. April 24 th happened to coincide with a national strike called by the opposition party, so Rana was expected to show that "his" workers were not on strike (Motlagh 2014: 64).
If the Rana Plaza collapse demonstrated in extremis the conditions of Bangladesh's export garment sector, so too does it manifest the social and spatial politics of labour rights in global supply chains. Under the international system established after the First World War, labour rights are governed as citizen-state relations, concerned with the regulation of employment relations and labour markets within a national territory. The ILO exemplifies this approach with the development of universal conventions that are territorialised by states through ratification and integration into national labour laws-a process that produces regulatory unevenness because not all countries adopt the same conventions (Alston 2004;McIntyre 2008;Standing 2008). Global supply chains transcend territorial nation-states and jurisdictions of labour law, permitting multinational corporations to profit from poor labour standards while evading liability. Garment-exporting countries are incentivised to suppress wages and labour standards, such that in Bangladesh many existing labour laws remain unenforced (Ahmed et al. 2014: 266-7).
With the inability or unwillingness of states to regulate factory standards, what has emerged over the past 30 years is a complex system of corporate self-regulation in the form of "codes of conduct" (Hughes et al. 2007, McGrath 2018. Even when based on ILO conventions, these codes are used to certify factories privately, usually by auditors paid for by apparel companies. A wide-ranging interdisciplinary literature speaks of the failures of voluntary codes for redressing the structural causes of poor labour practices (De Neve 2009; Egels-Zandén and Merk 2014; Ruwanpura and Wrigley 2011). By embracing codes of conduct, MNCs have been able to portray themselves as actively tackling labour abuses while opposing attempts to hold them legally liable for the conditions in supplier factories (Bartley 2005: 213;LeBaron and Lister 2015). In Bangladesh, codes of conduct deepen regulatory unevenness, with toptier suppliers strategically adopting and implementing codes to win contracts that bottom-tier suppliers cannot access (Tighe 2016). Before the collapse, Rana Plaza factories passed numerous independent certifications (Sinkovics et al. 2016: 625).
Geographers have demonstrated that neoliberal regulation is diverse, spatialised, and variegated (Brenner et al. 2010, Chari and Gidwani 2005, Werner 2016. As globalisation reconfigures labour rights, institutions like the ILO continue to prioritise the role of national governments, while workers "jump scale" with appeals to transnational anti-sweatshop campaigners and MNCs (Merk 2009). For this reason, scholarship on labour and global production has moved to also incorporate the diverse actors and institutions of neoliberal labour governance, including place-based analyses of labour rights contestations (De Neve 2008; Gunawardana and Biyanwila 2008;Kumar 2019;Padmanabhan 2011;Ruwanpura 2015). Transnational activists, working with local labour movements, mobilise the threat of boycott against companies accused of violating workers' rights (Seidman 2008). Analysing how rights are constituted in these spaces provides insight into how the unevenness referred to as "regulatory uneven development" is produced in the dynamic networks and institutions of neoliberal labour governance (Brenner et al. 2010;Werner 2016: 184-85).
Promoting labour standards through market mechanisms like boycotts and CSR privatises rights and narrows the range of labour struggles admitted into rights spaces. Successful campaigns tend to focus on the most egregious cases, such as child labour and modern slavery, because they carry a moral urgency that motivates the public, even if they are not workers' main concerns (Brooks 2007). This focus on extreme labour abuses is replicated in the ILO's elevation of four labour principles as universal human rights: the abolition of forced labour, the abolition of child labour, the elimination of employment discrimination, and the right to freedom of association and collective bargaining (ILO 1998;cf Alston 2004). An emphasis on extreme violations of "human rights"-conceived as universal, basic, and individual-can crowd out workers' "ordinary" claims for collective rights, which tend to pivot around routine labour issues like pay rates, working hours, and rights to occupational health and safety (McIntyre 2008;Spieler 2006).
Projecting place-based labour struggles into the marketplace of consumer choice and ethical consumption deterritorialises labour rights by prioritising consumers in the global North "caring" for distant others over the capacity of workers themselves to organize, form unions, bargain collectively, and strike (Anner 2012;De Neve 2008;Egels-Zandén and Merk 2014;Ruwanpura 2016). This reliance on sympathy and paternalism contributes to stereotypes of Northern consumers as "saviours" of poor workers, while failing to interrogate the relationship between low labour standards, cheap goods, and corporate profits (Siddiqi 2009).
Taken together, the privatising, narrowing, and deterritorialising of labour rights reflect neoliberal labour governance: the disembedding of historical labour rights from national contexts and their relocation into voluntaristic, transnational initiatives. This article focuses on how elite actors imagine, contend over, and reproduce privatised rights within transnational negotiations where MNCs wield formidable power. Workers' participation is often excluded from "global contestation" (Brooks 2007), and the Rana Plaza Arrangement was no different. Labour-centric analysis shows that workers have agency and advocate for themselves, and that place-based activism shapes economic geographies of capitalism (Padmanabhan 2011). Indeed, Bangladeshi garment workers challenge their circumstances with street-level protests, walkouts, and marches. Despite formal exclusion from sites of "global contestation," their periodic wins undeniably shape labour politics. How worker activism affected the Rana Plaza Arrangement in significant-though highly limited-ways is discussed below.
On the day of Rana Plaza's collapse, workers expressed concern about the building's safety, but went inside because they could not exercise a right to refuse unsafe work. In an environment of weak state protection, lack of organised labour power, voluntary regulation of MNCs, and a denial of labour rights, it may be surprising that global apparel companies paid $30 million compensation to workers in the aftermath of the Rana Plaza collapse, in a scheme that is widely described as "rights-based." Turning now to the demands and negotiations that produced the Rana Plaza Arrangement, we can see how the aim to channel resources to affected families eventually superseded calls for industry accountability. This language of "rights"privatised, narrowed, and deterritorialised-is critical to understanding the reproduction of regulatory unevenness in the post-Rana Plaza garment sector, where despite survivors' calls for both financial compensation and justice, some claims were more readily addressed than others.

From Calls for Accountability to Employment Injury Compensation
From the day of Rana Plaza's collapse, NGOs and government agencies provided medical and financial assistance to survivors alongside search and rescue efforts. Some apparel brands distributed cash and food parcels to affected families. Bangladesh labour law obligates factory employers to pay an extremely small amount of compensation after an occupational death or permanent disability, but these regulations are widely unenforced, and the amounts fall well below the international standards of employment injury compensation. 2 A private insurance scheme of the Bangladesh Garment Manufacturers and Exporters Association (BGMEA) that might have provided some compensation was restrictive, poorly managed, and not designed to provide payments on the scale of Rana Plaza (Miller 2012: 30, 53). Survivors of the collapse could sue their Bangladeshi employers for damages, but litigation would take years with no certainty that it would lead to a financial settlement.
There is no ready mechanism or institution for distributing compensation from multinational companies to survivors of global supply chain disasters. Bangladesh experienced repeated factory fires and collapses, but no large-scale compensation agreement with global apparel companies was reached until the 2005 collapse of the Spectrum sweater factory, in which 64 workers died (Miller 2012). That agreement, forged between Inditex (owner of the Zara fashion brand) and the textile trade union now known as IndustriALL, provided "financial relief" to injured workers and families of the deceased, paid for by the factory owner and some of the MNCs that sourced from Spectrum. The Spectrum compensation scheme was based on an adversarial industrial relations approach, where brands agreed to pay compensation for general damages like pain and suffering (rebranding compensation publicly as "relief") in exchange for claimants relinquishing their right to sue. As we will see, Rana Plaza survivors were compensated using a different approach, but the Spectrum relief fund proved it was possible for MNCs and labour groups to reach a settlement and administrate the complex process of defining criteria for eligibility, assessing needs based on level of injury, and implementing payments.
Based on their experience of Spectrum and other factory disasters, Bangladeshi and transnational labour rights groups believed that the best chance to get a fair distribution of what they called "meaningful" compensation to Rana Plaza survivors was to press the global apparel companies, government, and local employers to agree to a joint compensation deal while worldwide attention was still focused on the collapse. There are two main ways to achieve this: one, through civil lawsuits that might be won or resolved with a cash settlement; or two, by protesting against the brands until they were compelled to pay survivors even in the absence of a legal settlement. Civil lawsuits were indeed initiated against local employers and some global brands, but these were expected to take years and had no certainty of success (CCC and ILRF 2016). Hence, labour groups sought to mobilise public outrage to put money into the hands of survivors as soon as possible, sifting through paperwork in the rubble to identify brand names.
In addition to calling for compensation for the Rana Plaza families, garment workers, trade unionists, and bereaved family members at protests in Bangladesh demanded criminal punishment of those responsible-most prominently Sohel Rana, who fled the building collapse and was arrested at the Indian border (Motlagh 2014: 76). One Bangladeshi labour leader explained the approach of his NGO to me: "Our priority was compensation for the victims. We divided it into three phases: first rescue and rehabilitation, then to ensure compensation, and then justice with the criminal case so other [employers] should be cautious [in the future]. Both types of advocacy are correct-those who work for compensation and for criminal justice-but for us the priority was the compensation because people are starving, people are in a vulnerable situation. So our first responsibility was to ensure their protection." Transnational activist networks, including the Europe-based Clean Clothes Campaign (CCC), the U.S.-based Worker Rights Consortium (WRC), and the global trade union federation IndustriALL, used petitions, press releases, and direct actions at retail outlets to demand accountability from the global apparel companies. They called for a total of $71 million of compensation for the survivors' pain and suffering, lost wages, burial expenses, and medical needs. With dozens of brands and thousands of affected workers, the need for a single compensation agreement was explained to me by a representative of the Clean Clothes Campaign: "After Rana Plaza, the brands and retailers were just giving to various charities willy-nilly. It was an uncoordinated response, not based on rights. It was based on whether a charity thought you were deserving or not. Some people were getting lots of compensationespecially those that were 'photogenically injured'-and other victims who were less savvy were completed excluded." At the urging of these labour groups and an invitation from the Bangladesh government, the ILO agreed to serve as a neutral facilitator for bringing the MNCs and other "stakeholders" together to negotiate and deliver a compensation deal. The ILO had never been involved in compensation after an industrial disaster before, and it is a tribute to Rana Plaza's scale and gravity that an exception was made. As I explain below, the entry of the ILO into the compensation negotiations fundamentally shaped both the means and form of the final agreement, making it different from the compensation deal achieved after the Spectrum collapse. 3 The ILO hosted a meeting of Bangladeshi and global trade unions, labour rights NGOs, apparel companies, Bangladesh state officials, and Bangladesh employers' associations in September 2013. There and in subsequent, often contentious meetings, this group of representatives agreed broad principles of a "Rana Plaza Arrangement" in November 2013 with the signing of a Memorandum of Understanding (MoU) written by the ILO's legal counsel. Most of the signatories became members of the Rana Plaza Coordination Committee (RPCC), the Arrangement's governing body. 4 The extraordinary nature of this approach was highlighted by an ILO official who participated in the negotiations. Contrasting the collaborative and voluntary nature of the Arrangement to the legal settlements that were successfully negotiated by labour groups and MNCs after factory fires and collapses before, he said: "What had been done [after factory disasters] in the past was for the campaigners to go after the brands, and get what money they could to distribute to the victims. But no one before had set up the stakeholder engagement, a trust fund that received the funds to be donated, and to have a coordination committee that oversaw the implementation arrangements. That was all new." How did the ILO's role in brokering the compensation agreement and providing technical assistance influence the final deal? First, because the ILO has no legal grounds to enforce obligatory compensation, the Rana Plaza Arrangement had to be a voluntary agreement, not a legal settlement. The ILO established a trust fund into which companies could donate, and from which workers and bereaved families could "voluntarily" accept benefits without impacting their statutory rights (meaning that they could still sue factory employers or global brands).
A second consequence of the ILO's involvement was that the compensation deal was structured by the ILO's own institutional norms and standards. ILO actuaries and legal counsel promoted the use of ILO's Employment Injury Benefits Convention No. 121 (1964No. 121 ( /1980, which lays out compensation standards that would normally be implemented in a national employment injury insurance system. By modelling the Arrangement on the "universal" standards of ILO Convention No. 121, the Bangladesh government-which is yet to ratify the convention-would have a prototype that could be used later to establish a state system of social protection. Said one of the ILO actuaries: Convention No. 121 gave the Arrangement a basis in internationally-recognised labour rights, but meant that its payments were extremely limited: only covering lost wages, funeral expenses, and costs of medical care and rehabilitation. Importantly, the survivors would not be compensated for their "pain and suffering," because damages would usually be awarded by the court with a finding of liability, and are not part of Convention No. 121. The Rana Plaza Arrangement was therefore not a legal settlement to remedy the adverse human rights impacts of business operations (Hossain 2017;cf Ruggie 2008), but rather an agreement that instated workers' neglected rights to employment injury insurance. While the 2005 Spectrum relief fund aimed to be a replicable model for multinational corporations to compensate individuals affected by supply chain disasters in countries without robust employment injury insurance protection (Miller 2012: 142), the Rana Plaza Arrangement was intended to be a prototype for governments to set up an employment injury insurance system with the technical assistance of the ILO (Prentice 2018).
In January 2014, the ILO established a Rana Plaza Donors Trust Fund to collect donations and disburse payments. After months of vigorous campaigning by labour rights groups to urge companies to contribute, $30 million was raised for Rana Plaza's survivors, paid mostly by global apparel companies. 5 A Rana Plaza Claims Administration was set up in Bangladesh to accept and process the claims of workers and family members, which was completed in October 2015. A total of 5,109 claims were paid on behalf of 2,895 injured, deceased, or missing workers.
As a collective, non-adversarial, voluntary compensation deal, the Rana Plaza Arrangement did not broach questions of responsibility for the collapse. In this respect, there is a gap between the kind of settlement that the labour groups originally sought, and what they ended up with. As one UK-based activist put it: "We were stretched between these two goals: one was to get money for the victims who needed it, and on the other side, the stress was on trying to get this money from the people who were responsible." The MNCs who were needed to fund the compensation scheme (the Bangladesh government and local factory owners contributed comparatively little) were unwilling to accept any agreement that specified responsibility. Recognising that the Rana Plaza families could end up with nothing-failing to "get money for the victims who needed it"-labour groups accepted that the MNCs would contribute funds in whatever amount they chose, anonymously if they wished, without even disclosing whether they had sourced from Rana Plaza or not (several corporate contributors had no relationship to Rana Plaza). Although claimants would be entitled only to very basic amounts of compensation in line with employment injury insurance standards, these were substantially higher than what they would be entitled to under Bangladesh law (which effectively guaranteed nothing). With the MNCs willing to fund the scheme, labour groups-Bangladeshi NGOs and trade unions, and their transnational allies-accepted the agreement as a compromise.
What smoothed this compromise was the language of "labour rights." By using the standards of ILO Convention No. 121 to calculate the awards, the Arrangement recognised workers as bearers of rights that had been denied. This "rights-based" framework was enormously appealing to labour groups involved in negotiations, and made this deal different from the previous "relief" agreements between MNCs and survivors of previous factory disasters (Miller 2012). As a Clean Clothes Campaign activist working in Italy described it, approvingly: "We have to recognise the issue of compensation. It is not a donation, it is compensation. The strongest part of the [Rana Plaza] fund was that it was part of the ILO convention that recognises the workers' injuries." The quest for compensation was animated by two goals: to provide financial aid to survivors, and to hold accountable the institutions and actors who had failed in their duty of care. The ensuing agreement satisfied the first goal but not the second. MNCs funded the Arrangement on the condition that legal responsibility was removed from discussion. Punishment of industry actors-demanded by workers and their allies protesting in Dhaka-was no part of the Rana Plaza Arrangement because the ILO had no institutional means to adjudicate responsibility, and the scheme's funders (MNCs) were opposed to it. For this reason, Bangladeshi legal practitioners like Sara Hossain (2017) have questioned whether the Rana Plaza Arrangement can be rightfully called "compensation." Legal responsibility has been pursued separately-and for the most part, unsuccessfully-by the families of Rana Plaza's victims in the courts of Bangladesh, North America, and Europe (CCC and ILRF 2016).
Trade unions and labour rights NGOs-particularly Bangladesh-based organisations working closely with survivors-mourned the fact that by using only the ILO Convention No. 121 standards, there would be no payments that acknowledged the "pain and suffering" of Rana Plaza's victims. The $30 million settlement was less than half the amount the labour groups proposed, and only paid for lost wages and medical care. But with the language of "labour rights," these same activists could highlight the Arrangement's achievements in extending universal labour standards to Rana Plaza's survivors in a context where victories for workers are rare. They hoped, too, that the actuaries' careful design of the Arrangement in compliance with ILO Convention No. 121 meant that it could be used as a model for employment injury insurance, if the government of Bangladesh could be convinced to establish a state-led system. Although the government's poor record on labour rights made this unlikely (cf Bair et al. 2020), labour groups hoped that in the post-Rana Plaza period there might be a chance.

Privatising Labour Rights with Neoliberal Regulation
The Rana Plaza collapse dominated world headlines in April 2013, and it has returned to public consciousness periodically since. As Hasan Ashraf (2017) describes it, images of its aftermath-rubble piled high, bodies carried out on stretchers, bereaved kin clutching photographs, and everyday objects like ID cards, sandals, and paperwork nestled in the wreckage-have circulated in the news media and activist campaign materials as a recursive spectacle of suffering. Representing the deadly peril of a poorly regulated global industry, the Rana Plaza collapse is portrayed as a gross human rights violation (Siddiqui and Uddin 2016). For a $30 million compensation deal to emerge from these circumstances is historically unique but can be explained. The privatising of rights in the global garment industry means that big, photogenic violations of workers' rights receive attention from the general public, and action from labour rights advocates and the multinational companies they campaign against (cf Seidman 2008).
In the spatial politics of transnational labour organising, workers are essential to the imagery of labour rights violations, but rarely involved in decision making when it comes to solutions (Brooks 2007). Yet, labour geographers have demonstrated that despite workers' exclusion from elite negotiations, their activism transforms economic geographies in big and small ways (Padmanabhan 2011; Rainnie et al. 2011). An example of this is the mass protest movement of Bangladeshi garment workers that achieved a sectoral wage increase of 77 percent, seven months after the Rana Plaza disaster (Siddiqui and Uddin 2016: 694). In final negotiations over the Arrangement, labour activists in the RPCC-both non-Bangladeshi and Bangladeshi, trade unions and NGOs-successfully pressed for the compensation formula to reflect this higher amount, rather than workers' actual wages at the time of the collapse. So, while structural exclusion from "global contestations" narrows labour's influence, collective action can affect negotiations from the outside-albeit in limited ways-if labour groups use elite access to amplify grassroots activism.
As the final deal illustrates, the salient distinction is not only between who is "at the table" of elite negotiations and who is excluded. The differential power and leverage of those at the table, and the institutional environment in which negotiations take place are equally important. In the previous section, I showed how rights talk made compromise in the RPCC possible by granting labour activists a rare and highly public "win" for labour rights in a context where MNCs held the power of the purse strings. A transnational compensation agreement with the concept of rights so firmly at its centre invites reflection on how rights are conceptualised and operationalised in these kinds of voluntary initiatives.
The Rana Plaza Arrangement facilitated payments from multinational apparel companies to affected families, using a formula that defined rights based on what workers would have received from their direct employers under ILO Convention No.121 (ILO 2015;Prentice 2019). Applying the convention in this selective way is a common use of ILO Conventions in neoliberal labour governance, but not the purpose for which conventions were created (Standing 2008). The ILO sets universal labour standards for adoption and enforcement by countries in their national territories, but in neoliberal governance these standards are now frequently wrested from this context and repurposed to serve as free-floating "universal standards" of corporate self-regulation. A good example is the Ethical Trading Initiative's (ETI) "base code," which is a set of factory standards based on ILO core conventions, widely used in private factory audits (Barrientos and Smith 2007;Hughes et al. 2007).
Under ILO Convention No. 121, compensation payments for occupational injury or death are a percentage of the workers' wages at the time of the "accident," plus the medical costs arising from the injury (ILO 2015: 13). This standard is based on a historical formulation of social protection dating back to the nineteenth century, when employers' liability for worker injury was formulated as a safety net to prevent destitution and secondary health problems stemming from workplace harms (ILO 2014: 46-48). The limited aim of injury compensation justifies its scanty sum, on the expectation that a worker's wages are sufficient to begin with and need to be restored.
Geographers researching the global garment industry have demonstrated that geographies of uneven development are reproduced through the normalisation of diverse and novel forms of labour exploitation, with neoliberal regulation playing a crucial, orchestrating role (Bair and Werner 2015;Ruwanpura 2016;Werner 2016: 184-85). Examining the ILO's compensation standards within a framework of uneven development leads us to consider what it means to restore Rana Plaza workers to their previous conditions in a context where supply chains devalue labour across global terrain. Bangladeshi garment workers are paid poverty wages that place them on the edge of subsistence in the normal course of employment (Muhammad 2015: 147-148). In the RPCC, labour groups pushed for the use of higher wages and other fixes like a minimum payment floor to enhance the monetary value of the compensation awards. Still, by using the "safety net" standards of work injury compensation to calculate payments for Rana Plaza survivors-rather than, for example, what a court might award as damages for pain and suffering-compensation for death or permanent disability amounted to only a couple thousand dollars to replace a lifetime of lost earnings.
Drawing on her research among garment producers in the Caribbean, Marion Werner (2016) points to the neglected importance of history in the reproduction of uneven development. Though Werner emphasises the legacies of coloniality and racialised and gendered practices of subject-making in capitalist relations, in the Rana Plaza Arrangement it is the history of labour rights themselves that is apposite. Labour rights institutions-ranging from the tripartite ILO to the neoliberal governance of private codes of conduct-are constructed within histories of class struggle and class compromise (McIntyre 2008;Ruwanpura 2016;Selwyn 2012). The labour rights that are codified in ILO conventions are presented as universal rights for universal workers, but a critical labour perspective demands that we recognise the worker who serves as template for these rights is not the 21 st century Bangladeshi garment worker.
In the global North, labour rights like employment injury insurance were institutionally established in the twentieth century as part of a "grand bargain" between labour and capital: capital extends basic protections to workers, and workers accept capital's right to profit (Spieler 2017;Standing 2008: 365). ILO Convention No. 121 was developed within an integrated system to advance state guarantee of social protection with long-term disability benefits, unemployment benefits, and old-age pensions-not to mention labour standards pertaining to occupational health and safety to protect workers in the workplace, which Rana Plaza lacked. Applying ILO Convention No. 121 so selectively in the Rana Plaza Arrangement deforms the original aims of work injury compensation rights, which were formulated to function in coordination with other benefits.
In interviews with creators of the Arrangement, the ILO's indispensable role as broker was widely acknowledged. A common statement, in the words of one trade unionist, was: "without the ILO there wouldn't have been an Arrangement." But the institutional context also restricted the outcomes. Paying employment injury benefits to Rana Plaza's survivors appears to extend universal labour rights to workers who had been structurally denied those rights by state and industry actors. However, by treating Rana Plaza survivors as the worker-citizens of ILO conventions-simply deserving restoration of wages and payment of medical expenses after an industrial disaster-the Arrangement ignores the ongoing harm of the denial of rights in the first place. That MNCs and NGOs can trumpet their support of the Rana Plaza Arrangement as "rights-based" compensation adds to a feeling among some critics that the Arrangement has performed a sleight of hand: appearing to hold MNCs accountable while actually extending a self-congratulatory corporate power (Chowdhury 2017: 942). Given the enormity of the Rana Plaza disaster, the payment of basic employment injury compensation benefits can appear disproportionate to the harm, leading a chorus of activists, scholars, and survivors to insist that the Rana Plaza workers are yet to be properly compensated (cf Chen 2015, Chowdhury 2017, Muhammad 2015.

Conclusion
In its horrifying enormity, the Rana Plaza disaster motivated unprecedented responses from business, government, and civil society, often in the name of promoting or extending "labour rights" (ILO 2016). One such response was the Rana Plaza Arrangement, a $30 million voluntary compensation scheme facilitated by the ILO and funded mostly by global apparel companies. In a context where legal accountability for industry actors has been elusive, the Arrangement prioritised putting vital resources into the hands of affected families, and to do so in a way that respected "labour rights." Like the failed neoliberal regulations that it was designed to redress, the Arrangement is a hybrid that blends voluntarism and legal mandates, public and private, binding regulation and soft regulation, business and CSR (cf Bair 2017). By treating Rana Plaza's victims as the worker-citizens of "universal" labour standards, claimants received the wages that would have been paid had Rana Plaza not collapsed, but not additional compensation for the pain and suffering of those who survived or the families of deceased workers who did not.
Focusing on the social and spatial processes through which labour rights have become privatised, narrowed, and deterritorialised, this article contributes to our understanding of how rights are constituted in regimes of neoliberal labour governance. Bangladeshi garment workers-individually and collectively-struggle for improved conditions at work and in many forms of public protest (Ashraf and Prentice 2018;Siddiqi 2017). A transnational arena of globalised labour rights deterritorialises these struggles by projecting them into a privatised sphere of corporate self-governance. Much as Ethel Brooks (2007) has argued in relation to campaigns to boycott corporations over labour violations, the Rana Plaza Arrangement is an example of how the decision-making and imagination of "global" or elite actors shapes the labour rights terrain. If the Rana Plaza Arrangement began with Bangladeshi labour advocates and their international allies demanding that global apparel companies be held accountable by paying damages to families, it later became an opportunity for the companies to redefine their obligations after an industrial disaster by supporting the cause of "labour rights" with an exercise of corporate social responsibility.
The regulatory landscape of the global garment industry is diverse and variegated, but not randomly so (Ruwanpura 2016;Werner 2016: 184-85;cf Brenner et al. 2010). Regulatory unevenness can be reproduced through the enduring power of institutions, even when individual actors-such as the labour groups who participated in negotiating the Arrangement-try to achieve something different. This article's focus on the ILO as broker of the compensation deal shows the reproduction of regulatory unevenness in two important ways. First, despite public understandings of the Rana Plaza collapse as a violation of human rights, the compensation deal was not a legal settlement to satisfy survivors' right to remedy for human rights violations (Chowdhury 2017;Hossain 2017). Payments covered lost wages and medical care according to the ILO's labour rights standards, which would normally be overseen by the state as part of a wider social protection safety net.
The second way that existing regulatory unevenness shaped the outcome for workers is in the use of ILO Convention No. 121 to formulate compensation benefits. I have shown how the "universal worker" that serves as template for these benefits is historically and spatially specific, and not the Bangladeshi garment worker of the 21 st century supply chain. To examine the Arrangement in a framework of uneven development is to recognise that supply chains facilitate capital accumulation in places where labour rights are suppressed. With compensation payments that only cover lost wages and the costs of treating injuries, the Arrangement attempts to restore Rana Plaza's workers to their pre-collapse circumstances, where the very profitability of Bangladesh's export garment industry is premised on a strategic and ongoing suppression of labour rights. Despite the attempts of labour groups to extend labour rights to Rana Plaza's workers, the ILO's institutional norms and the strong bargaining position of the MNCs placed limits on which rights would be recognised, and how.
Most academic scholarship on Bangladesh's post-Rana Plaza garment industry focuses on novel transnational initiatives to prevent future disasters, such as the Accord on Fire and Building Safety in Bangladesh (cf Ashwin et al. 2020;Bair et al. 2020;Saxena 2020). A collaboration between global trade unions and multinational corporations to inspect and remediate garment factories, "the Accord" was originally developed after the 2005 collapse of the Spectrum factory, but only finalised in response to public outcry over Rana Plaza. A noteworthy feature of the Accord is that it makes the participating companies contractually liable for safety in the factories from which they source. Designed to be a transnational solution to the failures of national safety regulation, the Accord was widely criticised in Bangladesh for bypassing government institutions and intruding on national sovereignty, and has since been taken over by a local consortium that includes national employers' associations (Kang 2021;Saxena 2020).
The Rana Plaza Arrangement is also a transnational initiative, but because it is based on principles of social protection that normatively reside with the government, it invites an instantiation of state responsibility, not a circumvention of it. ILO actuaries always intended the Arrangement to be a prototype for a state-led system of social protection. Since the completion of the Arrangement in 2015, the ILO has been working with the government of Bangladesh and its employers' associations to design a national employment injury insurance scheme, to protect workers and employers from the financial threat of workplace "accidents" and avert the need for Arrangement-style agreements in the future (ILO 2016). The outcome of the project remains uncertain, amidst the concerns of factory employers and the government about the economic and political consequences of expanding social protection, and the reluctance of MNCs to make a long-term financial commitment (Russell 2018;cf Bair et al. 2020). 6 Civil lawsuits and the criminal prosecution of Rana Plaza's owner, factory managers, and labour inspectors are still making their way through Bangladesh's justice system (Ahamad and Islam 2019). Some Rana Plaza families filed lawsuits in Canada and the United States against the brands that sourced from the factories where their loved ones were killed, claiming negligence and wrongful death (Barrie 2016;Shaw 2017). These lawsuits have been unsuccessful-mostly because of statutes of limitation and plaintiffs' lack of standing-but they represent a growing strategy of transnational labour rights advocacy: to make enforceable across national jurisdictions MNCs' duty of care to workers manufacturing their products (Evans 2020). Long-existing barriers to successfully prosecuting global apparel companies for failures of due diligence are an enduring feature of global regulatory unevenness. Whether a guaranteed right to compensation for occupational injury or death will emerge from the wrongs of the Rana Plaza disaster hinges on the ability of labour advocates to convert the Arrangement into a more permanent solution, while developing judicial space for litigation against local and foreign actors for gross acts of negligence or harm.