The Rana Plaza disaster seven years on: Transnational experiments and perhaps a new treaty?

. This article examines several noteworthy initiatives that were implemented following the deadly 2013 Rana Plaza disaster in Bangladesh. They broke new ground in transnational labour law. The ILO-backed initiatives were largely successful but remain insufficient to achieve lasting change in the ready-made garment industry, where global brands’ supply chain buying practices constrain investment in occupational safety and health. A proposed United Nations treaty on business and human rights now seeks to enhance corporate accountability. Although promising, as part of a smart mix of multi-level public and private solutions, the treaty needs fine-tuning in the light of lessons learned from post-Rana Plaza experiments.


Introduction
The 2013 Rana Plaza disaster in Bangladesh, the world's second-largest readymade garment (RMG) producer, marked a watershed for business and human rights.Or did it?This article begins by tracing post-Rana Plaza developments, with a focus on responses involving the ILO.As significant as the results have been, they only scratch the surface of the root causes of unsafe working conditions in the garment industry, and of the challenges of obtaining remedies in case of harm.The fundamental asymmetries in the global value chain have meant that the three agreements that were signed in the immediate aftermath of the Rana Plaza disaster -the Accord on Fire and Building Safety in Bangladesh (the Accord), the Alliance for Bangladesh Worker Safety (the Alliance), and the Rana Plaza Arrangement (the Arrangement) -remain creative but incomplete transnational experiments.Given the continuing search for greater accountability in global value chains, one possible new avenue is the ongoing process in the United Nations (UN) towards the establishment of a binding instrument on business and human rights.This effort, which forms the analytical focus of the third section of the article, should draw on the lessons learned from the post-Rana Plaza experiments.The conclusion suggests that multiple responses are still needed in order to achieve greater respect for workers' rights, prevent similar tragedies in the future and bring those responsible for harm to account.

Post-Rana Plaza initiatives
2.1.The disaster and the changes that followed in Bangladesh On 24 April 2013, at least 1,132 men and women lost their lives when a building near Dhaka housing five factories collapsed, leaving more than 2,500 others injured in the rubble. 1 The garment workers inside the building had been manufacturing clothing for a range of foreign brands, under a series of commercial deals with suppliers and subcontractors.The day before the Rana Plaza building collapsed, cracks had been reported to an inspector, who ordered an evacuation.With pressing delivery deadlines, suppliers urged workers to return to work the next day or suffer dismissal (Kabeer, Huq and Sulaiman, 2019, p. 2).Those who went back ended up dead or severely injured.
In the aftermath of the disaster, surviving victims and the dependants of workers who had been killed began seeking justice.Criminal charges were brought against 38 individuals, but only a very few have ended up in prison (Ahamad and Islam, 2019;Alamgir and Banerjee, 2019, p. 284).In addition, a number of civil lawsuits have sought damages in the courts of various countries, but the cases have encountered many obstacles, such as arguments over jurisdiction, legal standing, causes of action, choice of law, statutes of limitations, liability theories, rules of evidence, measurement of damages and orders to cover costs (Salminen, 2018;Doorey, 2019; see also Kessedjian and Cantú Rivera, 2020).In particular, lawsuits have run up against "the conservatism of negligence law resistant to the imposition of a duty on corporations to protect supply chain workers from third-party harm" (Doorey, 2019, pp. 107). 2 While litigation can certainly put pressure on firms fearful of tarnished reputations, it is a slow, risky and expensive route.The proposed UN binding instrument on business and human rights (which I discuss in section 3) targets such obstacles and could perhaps be a "game changer".
In the meantime, those in search of justice looked for alternatives, such as the post-Rana Plaza transnational experiments examined below: the Accord, the Alliance and the Arrangement (with its Trust Fund), flanked by expanded technical cooperation programmes.The "international outrage over the incident" had laid bare the failure of brands' voluntary private initiatives (Berliner et al., 2015, p. 131; see also Bernaz, 2017;Martin, 2018).Bangladesh, where the RMG industry accounts for around 80 per cent of exports, attracted jobs by accepting the prevailing fast-fashion sourcing and purchasing practices of global brands.Under these practices, artificially cheap clothing for consumers has maintained downward pressure on working conditions and wages (ILO, 2016a;Human Rights Watch, 2019, pp. 2-5;Barrett, Baumann-Pauly and Gu, 2018).
With between 3.5 and 4 million people employed in the RMG industry in Bangladesh prior to the COVID-19 pandemic -60 to 80 per cent of them women -keeping such jobs in the country has been critical to Bangladesh's approach to development (Kabeer, Huq and Sulaiman, 2019;Rahman and Moazzem, 2017).Yet when suppliers pursue a low-price/low-wage strategy to attract business (Khan and Wichterich, 2015;Human Rights Watch, 2019), resistance to redistributive mechanisms such as collective bargaining should not come as a surprise.Nor should labour conflict, as seen in years of government repression of protests and strikes over trade union representation, wage levels and safety improvements (Rahman and Langford, 2012;Clean Clothes Campaign, 2020).The supervisory machinery of the ILO had long highlighted shortcomings in the Government's willingness and capacity to conduct inspections and to guarantee freedom of association. 3 The strong influence of RMG factory owners on the Government of Bangladesh, the fragmentation of trade unions, and their respective links to political parties, have complicated efforts to find a viable national response to the Rana Plaza tragedy (Alamgir and Banerjee, 2019, pp. 280 and 284;Rahman and Moazzem, 2017, pp. 83-84;Aizawa and Triparthi, 2015, p. 148).Instead, global buyers, international and local trade unions, non-governmental organizations (NGOs) and other actors have sought transnational solutions.

Accord on Fire and Building Safety in Bangladesh
Swiftly concluded by a wide range of stakeholders, the Accord broke new ground as a binding instrument that linked global market access to time-bound factory improvements, while taking on an ambitious and transparent programme of independent inspection, regular reporting, remediation, dispute resolution, training and law reform.The Accord was "unprecedented" due to the number of firms and workers involved, and to its legally binding nature (Reinecke and Donaghey, 2015, p. 257;Evans, 2015).With an initial lifespan of five years, extended by just over two years, both the original Accord of 2013 and its successor, the Transition Accord of 2018, linked market access through global brands to improvements by local suppliers.The Accord featured regular public reporting of aggregate inspection results and corrective action plans.It also exemplified the ILO's roles as convenor, neutral chair, broker and adviser.It illustrated the ways in which the ILO can move local parties beyond blaming each other to find solutions (Alvarez et al., 2019, p. 688).
The Accord achieved impressive results in the factories it covered, which were mostly first-tier suppliers, but it often did not reach further down the supply chain where hazards remained.The Accord's focus on fire and building safety, along with other occupational safety and health aspects, had to be flanked by other initiatives to strengthen worker voice and collective bargaining.While involving many parties, the Accord was hampered by the non-inclusion and then opposition of the influential local exporters' association.Overall, it was effective in meeting its immediate goals, but was not able to tackle the underlying causes of inadequate protection or the unintended blunt consequences of the disaster for local suppliers and workers.All the same, the Accord experiment offers valuable lessons that the current UN treaty process towards the establishment of a binding instrument on business and human rights should heed.

The conclusion and scope of the Accord
Immediately after the Rana Plaza building collapse, the global trade union bodies IndustriALL and UNIGLOBAL, along with 40 global garment buyers, asked the ILO to facilitate the negotiation of an agreement.The resultant Joint Statement on Building and Fire Safety (which led to the Accord) built upon the March 2013 National Tripartite Plan of Action on Fire Safety for the Ready-Made Garment Sector in Bangladesh (Diller, 2015, p. 336;Reinecke and Donaghey, 2015, p. 265).
The five-year Accord, which was finalized on 13 May 2013,4 eventually brought in 220 brands and retailers, the two aforementioned global trade union bodies and eight affiliated Bangladeshi trade unions.Four international NGOs witnessed the signature (Diller, 2015, p. 336;Reinecke and Donaghey, 2015, p. 274;Accord, 2018).The Accord aimed to provide independent inspections, remediation of hazardous fire and building safety conditions, prevention and redress (Diller, 2015, p. 336).The brands contributed (in weighted proportion to the annual share of RMG production in Bangladesh, subject to a cap) to covering inspection costs for their supplier factories, but not to funding remediation efforts.Under the Accord, suppliers were to respect workers' rights both to complain and to refuse unsafe work, without reprisals.It was also agreed that jobs would be protected while remediation was under way.The Accord also provided for safety training and worker participation.Yet this work was far from uncontested, since local exporters had not been included in the Accord's Steering Committee, out of fears of obstructionism.This later emerged as a major stumbling block for the Accord's continuation (Alamgir and Banerjee, 2019, pp. 288 and 291;Blasi and Bair, 2019, p. 18).5

The 2018 Transition Accord and transition to the RMG Sustainability Council
In June 2017, a group of global apparel firms and two global unions signed a second agreement to continue the factory safety programme in Bangladesh after the scheduled expiry of the Accord in May 2018. 6This pact, which became known as the 2018 Transition Accord, covered around 80 per cent of the factories included in the original Accord (Anner, 2018, p. 14), and was initially envisaged to run until 31 May 2021 and then until 2023 (Kang, 2019).The 2018 agreement retained features of the original Accord, extended mandatory joint occupational health and safety (OSH) committees to factories at all tier levels, and enhanced commitments in relation to respect for freedom of association (Accord, 2019a).While this was happening, however, the Bangladesh Garment Manufacturers and Exporters Association (BGMEA) decided to oppose the extension of the 2013 Accord, and supported a member's legal action, which led to a high court injunction against the Accord's operations in 2018 (Kang, 2019).After a series of events, an agreement between the Accord Steering Committee and the BGMEA, endorsed by the Government, eventually led to the resumption of these operations (Accord, 2019b).
In the meantime, preparations had begun for a transition from the Accord to a national mechanism with two pillars: the governmental Remediation Coordination Cell of the Ministry of Labour and Employment's Department of Inspection for Factories and Establishments (DIFE) and, from 31 May 2020, the private RMG Sustainability Council (RSC) (Accord, 2017(Accord, , 2018(Accord, , 2019b(Accord, and 2019c)).Under an agreement reached in January 2020 between the Accord and the BGMEA,7 the Accord's functions of inspection, remediation, training and complaint handling transitioned to the RSC on 1 June 2020 (Accord, 2020a).While also taking on local Accord assets and staff, the RSC is to operate on a voluntary basis (Accord, 2020a;Paton, 2020).
Despite certain guarantees (Accord, 2019b and2020b), there is considerable doubt about the scheme's viability and transparency (Clean Clothes Campaign, 2020;Paton, 2020).The RSC governance structure is to comprise representatives of the BGMEA and the corresponding knitwear association, global brands, and global and national trade unions, in cooperation with DIFE.It is unlikely, however, that this unit will be able to muster the same level of resources that were available for inspection under the Accord (some US$11 million per year, according to Anner, 2018, p. 3).
The RSC is to support DIFE in coordination with the ILO's Improving Conditions in the Ready-made Garment Sector Programme, which began six months after the Rana Plaza collapse and launched Better Work Bangladesh; its second phase is set for completion in 2023.Involving a range of stakeholders, including the tripartite actors, the Better Work programme entails collaboration between the ILO and the International Finance Corporation (IFC) to promote competitiveness in the garment industry by incentivizing compliance with international labour standards and overcoming obstacles to financing remediation (ILO and IFC, 2016;ILO, 2017).The programme has provided support for inspection, remediation, regulatory reform, creation of a safety culture and capacity building, including gender aspects (ILO, 2017 and2018).In an important parallel development, steps are being taken to bring a new national rights-based employment injury insurance scheme into effect as part of a sustainable social protection system (ILO, 2018).Furthermore, the ILO will be able to draw on the comments of its standards supervisory system in advising DIFE.

Results under the Accord
As an innovative transnational experiment with an emphasis on enforceability and transparency, the Accord represented complementary leveraging of capacities by labour and consumers to broaden the governance of the supply chain in the Bangladeshi RMG industry (Alamgir and Banerjee, 2019, p. 275;Evans, 2015;Reinecke and Donaghey, 2015, p. 267).The Accord has provided statistical updates on remediation, inspection, safety training, safety complaints and remedies (see, for example, Accord, 2020c), as well as quarterly aggregate reports (see, for example, Accord, 2020b).The disclosure of compliance assessments encouraged lead firms to source from compliant facilities (Blasi and Bair, 2019, p. 37).Among the key aspects of the binding Accord were the provisions for independent inspections, remediation, complaints and dispute resolution, and worker voice and training.

(i) Inspections and remediation
By January 2020, the Accord had completed many thousands of factory inspections in over 1,600 factories, with an initial remediation rate of over 90 per cent, with a further 400 inspections planned (Accord, 2020b).An earlier external study concluded that "the program has delivered an improved margin of safety for more than 2.5 million garment workers" (Anner, 2018, p. 1).This is no small feat.Another study of the health and safety situation found that factories under the Accord compared favourably to those covered by the National Tripartite Plan of Action, which tended to be lower down the value chain (Khan and Wichterich, 2015, pp. 33-36).A study involving 1,500 workers in the RMG industry concluded that "activities associated with [Accord] affiliation were either directly responsible for improvements in various aspects of working conditions or else they created a far more hospitable environment for other pressures to improve working conditions to have results than in non- [Accord] factories" (Kabeer, Huq and Sulaiman, 2019, p. 29).Yet much remains to be done, particularly in smaller factories lower down the value chain (Alamgir and Banerjee, 2019;Khan, 2018;Paton, 2020).The COVID-19 crisis now presents new health threats and the prospect of destitution for the RMG workers losing their jobs (Clean Clothes Campaign, 2020).
In terms of time-bound remedial measures under the Accord, of the 1,645 factories inspected over the relevant period, 156 had been given closure orders and a similar number had been relocated by January 2020 (Accord, 2020b).The Accord also resulted in temporary evacuations in at least 50 instances (Anner, 2018, p. 2).Factories that failed to implement the required safety measures were no longer eligible to source goods to Accord signatory brands (Accord, 2019a and2020a).The Accord's relative success has been attributed to requiring "lead firms to change their sourcing decisions in ways that make providing decent work in the economic interest of suppliers through both 'carrots' and 'sticks'" (Blasi and Bair, 2019, p. 6).
Yet the post-Accord prices paid by leading firms actually decreased (Anner, 2018), thus jeopardizing remediation funding and supplier support for the Accord.Shouldering the sometimes huge costs of remediation on their own, local firms argued that they were bearing the brunt of the low procurement prices paid by brands (Alamgir and Banerjee, 2019, p. 290;Khan and Wichterich, 2015, pp. 23-25).Delayed payments to suppliers did not help (Human Rights Watch, 2019).In July 2015, the IFC launched a credit facility specifically targeting remediation financing for RMG factories.While the IFC and similar initiatives fielded by other institutions freed up financial resources, a joint ILO-IFC study found difficulties such as high interest rates, a lack of financial literacy and bottlenecks (ILO and IFC, 2016).To this is now added the economic meltdown in the industry caused by the COVID-19 pandemic.If manufacturers are to be able to afford major improvements, this in turn implies changes in purchasing and pricing policies (Accord, 2018;Human Rights Watch, 2019).

(ii) Complaints and dispute resolution
Under the Accord, the most common complaints have thus far related to fire, structural and electrical hazards, health hazards (including working hours) and workplace violence (Accord, 2019a(Accord, , 2020b(Accord, and 2020c)).Collateral effects of the complaints mechanism were the reinstatement of some workers who had been fired illegally and the recovery of unpaid wages (Anner, 2018;Donaghey and Reinecke, 2018, p. 33).
The wider dispute resolution provisions under the Accord proved particularly important from a transnational law perspective.If a matter was not resolved by the Steering Committee within 21 days of the filing of a petition, the Accord established arbitration awards that became legally enforceable in the brand's home State, in line with specified international rules (para.5, 2013 Accord).In 2016 and 2017, the global union federations brought allegations before the Permanent Court of Arbitration, claiming that two companies had failed to require their suppliers to remediate facilities within the deadlines set, and had not negotiated commercial terms to make it financially feasible for their suppliers to do so.The Court rejected the respondents' admissibility objections, confirming the Accord's legal obligations and its enforcement mechanisms.8There was no decision on the merits, since the parties settled both claims (see Pietrogiovanni, 2018, pp. 274-275).In one of the cases, a US$2.3 million settlement was reached with a multinational apparel brand (Anner, 2018, p. 14).Even though the firms remain anonymous, the threat of arbitration has reportedly led to the resolution of a number of cases under the Accord (Blasi and Bair, 2019, p. 35).However, concerns over delays, costs, complicated procedures and other issues under the Accord's experimental dispute resolution system fed into what later became the Model Arbitration Clauses for Disputes Arising Under Enforceable Brand Agreements (Model Arbitration Clauses).9

(iii) Training and worker voice
By 2019, 1,200 OSH committees had received training, involving 1.8 million workers (Accord, 2019a).Women's lower literacy rates and comparative reluctance to make safety complaints led to the introduction of changes in OSH training (Alliance, 2014;Donaghey and Reinecke, 2018, p. 32).Experience gained through OSH committees helped to build worker participation (Kabeer, Huq and Sulaiman, 2019;Rahman and Moazzem, 2017).The Accord has won praise for its industrial democracy approach (Donaghey and Reinecke, 2018), with "the laborcorporate partnership" at the heart of the agreement (Accord, 2018).Some divergence in perception exists between global and certain local actors on this issue, with several smaller workers' organizations and NGOs feeling sidelined (Alamgir and Banerjee, 2019;Chowdhury, 2017).In any event, the gains won under the Accord are unlikely to become solidified under the new national mechanism without a longer-term entrenchment of respect for freedom of association and the right to engage in collective bargaining -in law, practice and attitudes (ILO, 2016a, pp. 13-14).A combination of collective bargaining and collective brand action could be especially effective (Human Rights Watch, 2019, pp. 29-30).

The Alliance for Bangladesh Worker Safety
In addition to the Accord, a second transnational initiative emerged in the North American-led Alliance for Bangladesh Worker Safety (the Alliance). 10In contrast to the Accord, the Alliance initiative involved the business side only, imposing no binding force or job guarantees (Aizawa and Triparthi, 2015, p. 146).The Alliance constituted a five-year commitment by 27 (later 29) retailers to support Alliance participants to improve building and fire safety in 587 factories (Donaghey and Reinecke, 2018, p. 25;Khan and Wichterich, 2015, p. 37).The governance structure included the BGMEA, but its president withdrew the Association in 2015.
The Alliance included brand-financed inspections and safety training, the provision of loans, and the means to channel contributions to the compensation fund (see section 2.4.1 below).In addition to the inspections and remediation conducted under the Alliance, its work contributed to building the knowledge base.But the Alliance did little to develop organizational capacity and basically involved "brand benevolence" rather than the Accord's "labour-negotiation" approach (Donaghey and Reinecke, 2018, p. 30).When it came to payments to victims of the collapse, or to surviving dependants, companies in both the Alliance and the Accord were involved in the Arrangement (see section 2.4.1 below).
The Alliance stopped work in December 2018.Its successor, Nirapon, involves 23 of the Alliance brands (Glover, 2019).It may provide inspections, training and advice, but has no mandate to rate, order remediation in, or suspend factories.In December 2019, Nirapon lost an appeal against an earlier court decision that had imposed a six-month ban on inspections (ibid.).In a sense, the less ambitious Alliance prepared the ground for what has become the more business-driven successor to the Accord, the RSC.

The Arrangement
The Rana Plaza disaster left the victims' survivors and the many injured workers clamouring for compensation for the harm that they had suffered.However, Bangladesh lacked an effective mechanism to provide compensation for the dependants of deceased workers and for surviving workers in need of long-term medical care (ILO, 2015).Lump-sum payments would have been very small and not up to minimum standards (ILO, 2018; see also Prentice, 2019).In terms of effective redress, there was clearly a void -the very type of void that the proposed UN treaty purports to remedy (as discussed in section 3).
In September 2013, the Clean Clothes Campaign and the global union federation IndustriALL put the problem to the ILO, which brokered multi-party consultations.Labour groups proposed a US$74 million compensation scheme that would have covered long-term losses in earnings and payment for pain and suffering (Prentice, 2019).However, most of the global apparel firms, which were not the direct employers, rejected any legal liability, avoiding the terms "compensation" and "rights".They were instead willing to provide "charitable" contributions for distribution among the victims (ibid.).The compromise "Understanding for a Practical Arrangement on Payments to the Victims of the Rana Plaza Accident and their Families and Dependents for their Losses" (the Arrangement) was concluded on 20 November 2013.11The parties to the Arrangement were the Government of Bangladesh, the BGMEA, the Bangladesh Employers Federation, several global brands/retailers, global and local trade unions, and an international and a local NGO (ibid.). 12The text of the Arrangement included "Practical Guidelines" that aimed to provide victims with payments in "a transparent and equitable manner" (Arrangement, para.1(a)).The multi-stakeholder Rana Plaza Coordination Committee (RPCC) was tasked with implementing the Guidelines.The ILO's role in relation to the Arrangement was, as in the case of the Accord, to act as a neutral and independent chair and to provide technical expertise upon request, without direct responsibility for implementation (para.5).
A key feature involved following a "single approach" to claim handling, "consistent with" ILO standards (para. 1 of the text's Practical Guidelines), in particular the principles of the ILO Employment Injury Benefits Convention, 1964 (No. 121), even though Bangladesh had not ratified it.The RPCC established methods for benefit calculation and distribution, set up an expert Claims Administration, and designated local NGOs to provide pre-claim counselling and assistance to victims.A worker permanently injured in the Rana Plaza disaster would receive lifelong monthly payments of 60 per cent of the wage payable in October 2013 (thus benefiting from a post-disaster wage increase), with a percentage factoring in the person's level of disability.A formula was also devised for lump-sum payments to dependants of deceased workers, taking into account the number of dependants to be covered by each claim and the relevant provisions of Bangladeshi law (Prentice, 2019).Actuarial assistance supported all calculations.
The formulae used for compensation were the result of a compromise backed by technical competence and an international labour standard.Reflecting national employment injury schemes that arose in the early twentieth century as an alternative to fault-based negligence tort law, under these provisions the Rana Plaza victims consented to the final and exclusive nature of payments. 13n addition, victims who were left too injured or traumatized to return to the garment industry also received training, counselling and medical support through the Arrangement (ILO, 2017, pp. 5, 28, 30 and 31).

Rana Plaza Trust Fund
At the request of the parties to the Arrangement, in January 2014 the ILO set up the independent Rana Plaza Trust Fund to receive third-party donations for victims, collected principally through the Accord and the Alliance.14Labour advocates and civil society groups launched public campaigns to pressure brands to donate (Blasi and Bair, 2019, p. 28).International clothing brands contributed the majority but not all of the funds collected (a few governments also donated).The slow pace of contributions led to some frustration among claimants (Prentice, 2019).Some firms contributed to NGOs operating outside the RPCC framework (Chowdhury, 2017).
The first claim submissions under the Arrangement were received 11 months after the building collapsed.By October 2015, with the ILO as trustee, the Trust Fund had disbursed over US$30 million to 5,109 beneficiaries -an unprecedented result.A further sum exceeding US$1 million has covered ongoing medical costs through the Trust for Injured Workers' Medical Care including for Rana Plaza workers, established in October 2016 (with government representatives, the employers' federation and trade unions serving as trustees).Workers with more serious permanent disabilities have benefited from an arrangement made with a local NGO.Three claims commissioners (one international and two national) determined the amount of compensation to be awarded to each claimant within the agreed framework.Payment delivery has been monitored to ensure the meaningful use of funds and to safeguard against fraud.The Arrangement and the Trust Fund saw themselves as performing "bridging" functions towards a longer-term provision at the national level -the still-needed employment injury insurance system (Clean Clothes Campaign, 2020).

Assessment of the Arrangement and the Trust Fund
The Arrangement's self-portrayal seems justified: "a ground-breaking initiative, which, for the first time, brought together representatives from along the supply chain to agree to a comprehensive and independent process that would deliver loss of income payments to those families impacted by the Rana Plaza disaster". 15 And the total compensation involved was significant.Concerns expressed about the scheme's "whitewashing effect" for consumers, regulators and adjudicators (Salminen, 2018) seem exaggerated.They ignore the reality that had each individual worker or his/her family been left to pursue compensation in the local courts, they would still be waiting and would perhaps be out of pocket due to legal fees.Injured workers and the families of those who perished received payments relatively quickly, under a single approach.
However, treating the matter as one of charitable contributions essentially perpetuated a denial of firms' responsibility for the potential costs of doing business.In any event, the methodology developed for compensation after the disaster, based on international labour standards, can and has been used in other cases, such as the Tazreen Claims Administration Trust and the Ali Enterprise compensation scheme in Pakistan (Prentice, 2018).Thus it may be that the post-Rana Plaza initiatives with the most lasting impact will turn out to be the Arrangement and the Trust Fund, rather than the more context-specific Accord.

Overall assessment of the post-Rana Plaza initiatives
Overall, the combined efforts of the Accord, the Alliance, the Arrangement/ Trust Fund and the ILO/IFC Better Work Programme in Bangladesh have had results: some rapid payments and other measures for victims; the improvement of structural, electrical and fire safety in most of the RMG factories covered; strengthened local inspection capacity; the resolution of numerous complaints; provision of training; and new avenues for participation that have also strengthened worker voice.As the Permanent Court of Arbitration procedural rulings and ensuing settlements showed, the path-breaking Accord created legally binding obligations that were ultimately largely respected (despite being contested, time-limited and non-universal in the RMG sector).The experience fed into the new Model Arbitration Clauses, which also drew on The Hague Rules on Business and Human Rights Arbitration of 2019 16 and United Nations Commission on International Trade Law (UNCITRAL) Arbitration Rules (as revised in 2013). 17Under the Arrangement, while the compensation scheme also attracted some criticism, the ILO, working creatively with partners, brought representatives of diverse interests to the table to find internationally inspired, practical solutions to get money into the hands of survivors.The ILO's exercise of its roles of convener, neutral broker, trustee and expert adviser in response to the Rana Plaza 15 See "Introduction", at: https://ranaplaza-arrangement.org. 16 disaster has legitimated the Organization's value.However, with 187 member States, most of which are developing countries, ILO resource levels would not permit replication of a response on this scale every time such a need arose.
All the same, the lessons learned about transparent governance, design and operations under these initiatives (ILO, 2015) could also refine current efforts to adopt new international rules on access to remedies (see section 3).As well as illustrating the useful role that the ILO can play, the post-Rana Plaza experiments offered lessons on the importance of transparency (such as the disclosure of compliance assessments) and enforceability (of arbitration awards, for example).The streamlined approach to handling independent claims, building on technical expertise and international labour standards, could also be emulated elsewhere.Learning from the post-Rana Plaza initiatives in relation to industrial democracy and worker participation, future instruments should emphasize respect for freedom of association and collective bargaining, and be gender-responsive.
While time-bound experiments and programmes can have important demonstrative effects, they do not in themselves change fundamental asymmetries along the value chain.They "do little to address the sourcing, pricing and procurement practices in the garment industry, which are the key reasons for the deplorable conditions of workers and that remain outside the mandate of these compliance mechanisms" (Alamgir and Banerjee, 2019, p. 293; see also Berliner et al., 2015, p. 142).When factories continue to base their manufacturing strategies solely on very low wages and over-extended production capacity, the resulting long working hours lead to efficiency, quality and safety issues (ILO, 2017, p. 25;ILO, 2016a).Local suppliers argue that the pricing policies of global brands offer no alternative (Human Rights Watch, 2019; Paton, 2020).Indeed, placing obligations for remedying deficiencies solely at the level of the production country seems inadequate, since the global brands, importing countries and, indeed, their consumers are benefiting from the status quo (Locke, 2013;Human Rights Watch, 2019).New rules favouring greater accountability in cases of harm from business activities would contribute to altering that paradigm (Bernaz, 2017;Simons, 2014).

A potential new treaty on business and human rights
The sheer scale of the Rana Plaza disaster indirectly spurred developments in a number of other areas, from national legislation 18 to regional initiatives (see, for example, Smit et al., 2020), ILO resolutions, 19 and what is known as the "UN 18 In France, for example, Act No. 2017-399 of 2017 concerning the duty of care of parent companies and contracting enterprises (Cossart, Chaplier and Beau de Lomenie, 2017).For a range of examples of measures, see ILO, 2016b, pp.41-61 and United Nations, 2018, pp.16-20.

Background to the proposed treaty
In 2011, the UN Human Rights Council adopted the UN Guiding Principles on Business and Human Rights (UNGPs). 20The non-binding UNGPs are based on three pillars: the State's role to protect human rights, business' role to respect human rights and, as a joint responsibility, access to remedy for victims in case of business-related human rights violations (United Nations, 2020). 21Firms are to engage in due diligence (see especially Principles 15 and 17-21), and to cooperate in remediation where they have caused or contributed to adverse human rights impacts (Principle 22).Although the UNGPs have attracted criticism (see, for example, Thompson, 2017, p. 57;Trebilcock, 2015, pp. 94-96), they have emerged as the dominant framework on business and human rights (see, for example, ILO, 2016b;OECD, 2018). 22In the labour context, serious due diligence can imply a substantial investment (Trebilcock, 2015;McCorquodale et al., 2017;United Nations, 2018).Guiding Principle 25 recalls the duty of States to ensure access to effective remedy as part of their duty to protect, and Principle 26 urges a reduction of legal, practical and other barriers to access remedies.While Guiding Principles 28, 30 and 31 relate to non-State-based grievance mechanisms (including multi-stakeholder and other collaborative initiatives), recent UN guidance on the use of the UNGPs has made little or no mention of grievance mechanisms under collective bargaining agreements, labour arbitration or international framework agreements as possible avenues for access to effective remedies (not to mention the Rana Plaza experiments). 23This stands in contrast to the ILO Resolution concerning decent work in global supply chains of 2016, which referred to them while also calling for examination of which measures, initiatives or standards are needed to promote decent work in this area (paras 11, 17-18 and 23(c)). 20See United Nations: Guiding Principles on Business and Human Rights: Implementing the United Nations "Protect, Respect and Remedy" Framework (A/HRC/17/31) and Human rights and transnational corporations and other business enterprises (A/HRC/RES/17/4), General Assembly, 63rd Session, New York, NY, 2011.Explained in United Nations, 2020. 21The labour rights dimensions of the UNGPs derive from their reference to the ILO Declaration on Fundamental Principles and Rights at Work (1998) and the two UN Covenants of 1966, in particular Art.7 of the UN Covenant on Economic, Social and Cultural Rights guaranteeing "safe and healthy working conditions".See also United Nations: General Comment No. 24 (2017) on state obligations under the International Covenant on Economic, Social and Cultural Rights in the context of business activities, Committee on Economic, Social and Cultural Rights, E/C.12/GC/24, New York, NY, 2017. 22See also ILO: Resolution concerning decent work in global supply chains, 2016, op.cit. 23The UN Working Group on the issue of human rights and transnational corporations and other business enterprises has mentioned "providing national labour tribunals with the competency to adjudicate business and human rights impacts" (United Nations, 2014, p. 29; see also United Nations, 2017a, p. 21), but this language was dropped from the final version of guidance on national action plans under the UNGPs (United Nations, 2016).In any event, how would such bodies be able to deal with complex tort or commercial contract litigation following an industrial disaster?
Dissatisfied with the voluntary approach of the UNGPs, several developing countries pushed for the adoption of UN Human Rights Council Resolution 26/9, which in 2014 tasked the Open-ended Intergovernmental Working Group (OEIGWG) with elaborating "an international legally binding instrument on transnational corporations and other business enterprises with respect to human rights". 24Under Ecuador's leadership, treaty proponents have moved ahead in the face of substantial opposition from many governments and the business community (Deva and Bilchitz, 2017).The Chair of the OEIGWG released the preliminary draft of a legally binding instrument in July 2018, which was discussed by the Working Group in 2018 and 2019. 25The initial set of proposals included attempts to impose direct legal obligations on corporations (United Nations, 2017b), but this was then abandoned.More recent drafts would turn many of the nonbinding features of the UNGPs into obligations for ratifying States.While engagement in the negotiations has increased, a number of States continue to oppose this translation of "soft" law into "hard" law.

Proposed content of the draft instrument
In July 2019, the Chair of the OEIGWG issued a revised draft of a "Legally binding instrument to regulate, in international human rights law, the activities of transnational corporations and other business enterprises",26 which was discussed in October 2019.On 6 August 2020, the Chair published a new version for examination by the OEIGWG in October 2020.27If adopted and implemented (big "ifs"), this treaty could indeed be a game changer for accountability in global supply chains.Without going into detail, we will now look at the most relevant provisions of the most recent draft from a post-Rana Plaza perspective.

Purpose, scope and definitions
The purpose of the instrument (draft Art. 2) focuses on state obligations to respect, protect and promote human rights in the context of business activities, including the prevention of human rights abuses (draft Art.6), ensuring effective access to justice and remedy for victims (chiefly draft Arts 4 and 7), and facilitating and strengthening mutual legal assistance and international cooperation (draft Arts 12 and 13).The draft's scope is capacious: it covers a broad range of human rights and all business activities (draft Art. 3). 28It arrives at this sweeping 24 United Nations: Elaboration of an international legally binding instrument on transnational corporations and other business enterprises with respect to human rights, General Assembly, 26th Session, A/HRC/RES/26/9, New York, NY, 2014.approach through a series of definitions (in draft Art.1), including of "business relationship" (draft Art.1(5)).29 Through this approach, the instrument would essentially regulate accountability along supply chains from the viewpoint of human rights -a change in paradigm from a business law logic, particularly in the RMG sector.At the same time, however, the latest draft reflects the UNGP approach of permitting state laws to differentiate how business enterprises discharge their obligations "commensurate with their size, sector, operational context and the severity of impacts on human rights" (draft Art.3(2); see also draft Art.6( 2)).This may attenuate some opposition to the proposals.

Due diligence and prevention
Effective regulation of the activities of business enterprises and human rights due diligence requirements lie at the heart of the text on prevention (draft Art. 6).The due diligence to be undertaken by business enterprises in relation to identification and assessment, prevention, mitigation, monitoring and communication explicitly extends to human rights abuses (the term "violations" found in the July 2019 draft has been dropped) that may arise from business relationships (draft Art.6 (2)).The draft calls for regular human rights impact assessments, integration of a gender perspective at all stages of the due diligence process, inclusion of due diligence requirements in contracts, and non-financial reporting, including on policies, risks, outcomes and indicators on labour rights throughout an enterprise's operations, including in their business relationships (draft Art.6(3)(a), (b), (e) and (f)), among other measures.

Remedies and redress (i) Victims' rights, access to justice and remedies
Draft Article 4 lists the rights of victims, while draft Article 5 expands upon their protection as they seek a remedy, and draft Article 7 concerns access to remedy.Victims' rights include the right to fair, adequate, prompt and nondiscriminatory access to justice and effective remedy (Art.4(2)(c), and the right to submit claims, including through class action, to courts and state-based nonjudicial grievance mechanisms (Art.4(2)(d)).The remedies mentioned include "restitution, compensation, rehabilitation, satisfaction [and] guarantees of nonrepetition" for victims (draft Art.4(2(c)).This is essentially what Rana Plaza victims were seeking, and to some extent received, under the Arrangement or the Accord.

(ii) Legal liability
In relation to legal liability for human rights violations or abuses in the context of business activities (draft Art.8), States are to adopt measures to ensure that their domestic jurisdiction "provides for adequate, prompt, effective, and gender responsive reparations to the victims of human rights abuses in the context of business activities" (draft Art.8( 5)).Governments are to make natural or legal persons conducting business subject to liability for: failure to prevent another legal or natural person with whom it has a business relationship, from causing or contributing to human rights abuses, when the former legally or factually controls or supervises such person or the relevant activity that caused or contributed to the human rights abuse, or should have foreseen risks of human rights abuses in the conduct of their business activities, including those of transnational character, or in their business relationships, but failed to put adequate measures [in place] to prevent the abuse (draft Art.8( 7)).
This far-reaching proposal remains controversial and could imply changes in contracting and pricing practices (see, for example, OECD, 2018;Human Rights Watch, 2019).It adds that courts are to decide on liability after examining compliance with "applicable human rights due diligence standards", but that such due diligence will not automatically absolve a legal or natural person from liability (draft Art.8( 8)).The text also provides for the possible reversal of the burden of proof, "consistent with the rule of law requirements" (draft Art.7(6)).

(iii) Financial aspects and assistance
In some jurisdictions, a court may order unsuccessful plaintiffs to reimburse the legal expenses of the other party,30 which may act as a powerful deterrent to victims' pursuit of remedy.The July 2019 draft precluded this as well as possible requirements to furnish a warranty to commence proceedings (in its Art.4(12)-( 13)), while the more recent version does not.The instrument provides for an International Fund for Victims, but without giving particulars (draft Arts 13(2)(e) and 15(7)).It is unclear how a generalized international fund would enhance the accountability of particular firms in industries, such as RMG production, which are not inherently hazardous.Draft Articles 12 and 13 call for mutual legal assistance and international cooperation.Such measures would have been helpful for some Rana Plaza complainants.

(iv) Criminal responsibility
The text also contains, in draft Articles 8( 9) to (11), stipulations on criminal liability.Article 6 of the July 2019 draft had listed as criminal offences, inter alia, forced labour as defined in the Forced Labour Convention, 1930 (No. 29), and the Abolition of Forced Labour Convention, 1957 (No. 105), but without mentioning the Protocol of 2014 to the Forced Labour Convention, 1930 (which itself refers to adequate remedies) and the use of child soldiers, as defined in the Worst Forms of Child Labour Convention, 1999 (No. 182).The more recent version drops these specific references.However, where such acts occur, along with those relating to denial of freedom of association or discrimination that are of a criminal nature, the scope of the provisions regarding "fundamental ILO Conventions" (draft Art.3(3)) could still capture them.

Overcoming hurdles
The draft treaty also contains provisions on adjudicative jurisdiction over claims, a reasonable statute of limitations for civil law claims, and choice of applicable law (draft Arts 9, 10(2) and 11) -the very dilemmas that supply chain litigation has faced in various jurisdictions following the Rana Plaza disaster.The draft sets out the jurisdictions in which claims may be brought (see details in draft Art. 9) and it would preclude courts from declining jurisdiction based on the doctrine of forum non conveniens (draft Art.9(3)), which will not be easy for some States to accept.
Moreover, States are to provide "effective mechanisms for the enforcement of remedies for human rights abuses, including through prompt execution of national or foreign judgements or awards, in accordance with the present (Legally Binding Instrument), domestic law and international legal obligations" (draft Art.7( 7)).The text contains details on the facilitation of recognition and enforcement of judgments (draft Art.12( 8) and ( 9)).Such draft Articles, along with those on legal liability, would make it easier for victims of workplace disasters to pursue claims and have court judgments honoured.On the other hand, the provisions on consistency with international law principles and instruments (draft Art.14) contain language on territorial integrity and non-intervention in the domestic affairs of other States that could blunt the impact of the treaty as a whole.To avoid this pitfall, more attention should be given to alternative means of dispute resolution such as streamlined binding non-judicial arbitration (for example, the Model Arbitration Clauses), which should also be more affordable than litigation.

Institutional arrangements
The draft puts forward institutional arrangements that include a traditional committee of human rights experts and a conference of State Parties (draft Art.15(1) to ( 6)).It would permit a State to accept the compulsory submission of a dispute to the International Court of Justice or to arbitration as mutually agreed (draft Art.18).In this connection, why not provide a role for the ILO in establishing a list of international arbitrators qualified to address labour issues (see Blasi and Bair, 2019, p. 36), or, with ILO consent, provide for possible referral by the instrument's committee of experts to ILO supervisory mechanisms when appropriate?

Draft optional protocol
The draft treaty would be open to protocols (draft Art.17).In September 2018, the OEIGWG put forward a draft optional protocol to the proposed treaty.31Under it, a State would need to designate or establish a national implementation mechanism to promote compliance with, monitor and implement the new treaty.The optional protocol would also empower the committee envisaged under the treaty to consider individual and collective communications alleging violations (draft Arts 8 to 12, Optional Protocol).

What is the added value of the draft treaty and protocol?
The proponents of the treaty see it as complementary to a better and more nuanced notion of business and human rights and of the implementation of genuine protection and remedy for victims (Baxi, 2016;Deva and Bilchitz, 2017).It would arguably strengthen States' duties to protect human rights and would impose duties of mutual legal assistance on States to ensure access to effective remedies for victims (De Schutter, 2016, pp. 44-55 and 63-66).And while international labour Conventions include measures to give them effect (United Nations, 2015), some are silent or vague in relation to remedies for victims.In this and other ways, the draft treaty could provide some added value.Yet recalling that the impact of the UNGPs was inherently linked to their non-binding nature, some opponents fear that such a (binding) treaty could "hinder or weaken existing guarantees and their natural evolution" (Černič and Carrillo-Santarelli, 2018, p. 299).For others, that evolution is simply too slow.The treaty process could help to identify and bridge the gaps that limit effective access to remedy.
Still, both the draft treaty and the draft protocol leave a lot of legal questions unanswered, such as, what is their relationship to the newly adopted Convention on the Recognition and Enforcement of Foreign Judgments in Civil or Commercial Matters of 2019, 32 and the Convention on International Access to Justice of 1980? 33In general, the treaty drafting process has, so far, not reflected sufficient awareness of these or other existing international legal obligations.For instance, Article 14(5)(b) of the draft Legally Binding Instrument states that any new agreement on trade and investment "shall be compatible with the States Parties' human rights obligations" under the new treaty.Yet how would this provision fit in with the general international law rule giving precedence to subsequently concluded treaties? 34 The proposed instrument requires further legal fine-tuning, along with more appreciation of alternative forms of dispute resolution in the field of labour and greater acknowledgement of relevant ILO instruments and mechanisms (as in United Nations, 2015 and 2016).Drafters could also be more imaginative in relation to the monitoring provided under the potential new instruments, establishing links to ILO bodies where ratified ILO Conventions or ILO principles are involved.The governance lessons of the post-Rana Plaza experiments should be taken into account to strengthen the language on prevention, dispute resolution and remedies.
In any case, the treaty process cannot escape the fact that it is captive to a statist model, from adoption to ratification.This entails two challenges.The 32 Available at: https://assets.hcch.net/docs/806e290e-bbd8-413d-b15e-8e3e1bf1496d.pdf.This Convention, adopted unanimously by the Hague Conference on Private International Law on 2 July 2019 would, upon entry into force, complement the Hague Convention on choice of court agreements of 2005, and in some measure the Convention on the recognition and enforcement of foreign arbitral awards of 1958.The Accord on Fire and Building Safety in Bangladesh referred to the 1958 Convention. 33Available at: https://assets.hcch.net/docs/a311a685-d6e7-41d4-8210-7c2b8c30429e.pdf. 34Art.30 of the Vienna Convention on the Law of Treaties of 1969 (newly mentioned in draft Art.14(4) on state immunity) sets out the means of applying successive treaties relating to the same subject matter.
first is that while civil society exerts considerable pressure in the UN context, States take the decisions.Second, would countries that fail to respect existing labour and other human rights instruments do better under a new instrument that calls for remediation in the event of their violation?The proposed treaty only skirts around the more fundamental predicament of transnational labour law, namely that globally generated problems are constantly pushed back down to be resolved at the national level (see Blackett, 2019;Pekdemir, Glasbergen and Cörvers, 2015).The draft treaty relies heavily on domestic legislation for implementation.
Nonetheless, such a new instrument would increase the pressure on firms to conduct serious human rights due diligence to prevent harm and be held accountable in case of failure.Even when not ratified, the treaty would at least provide ammunition for those calling for a range of measures to be taken with regard to social justice in global value chains (see, for example, Anner, Bair and Blasi, 2013;Bernaz, 2017;Hyde, 2016;Simons, 2014).In addition, it might give impetus to other suggestions, such as requiring that firms' financial plans include provisions for the possible costs of clean-up and medical care, should environmental and social harm result from their activities (ILO, 2019, p. 50).This would, like the Accord, combine market mechanisms with legal obligations of transparency and accountability.

Conclusion
Most of the value from the garment manufacturing supply chain is captured by brands and retailers, with little left for suppliers and still less for workers (Human Rights Watch, 2019;Alamgir and Banerjee, 2019;Lund-Thomsen and Lindgreen, 2014).The basic dilemma of global value chains remains: "a high level of integration between manufacturers and the brands and retailers that source their product from … far-flung vendors, combined with a legal regime that absolves those who command the various links in the supply chain of the kind of responsibility, moral as well as economic, that attaches to those in formal leadership of the vertically structured corporation" (Lichtenstein, 2017, p. 330; see also Bernaz, 2017;Martin, 2018).With their retail stores shuttered by COVID-19 restrictions, numerous buyers have cancelled contracts with suppliers, reminding us of the volatility of the RMG industry and the fragility of an economy and a workforce dependent upon it.These are problems which the Rana Plaza experiments were not designed to tackle head-on, but they open up space for deeper reflection about how to construct global value chains more equitably and sustainably.Once again, the ILO is convening major players to find a path out of a crisis in the Bangladeshi RMG industry; what is needed more broadly is a "new normal" (United Nations, 2020).
The experiments of the Accord, the Alliance, and the Arrangement/Trust Fund offer insights that should be taken into account in designing future approaches, from the local to the transnational.The post-Rana Plaza initiatives have advanced transnational labour law by moving it onto new terrain to reach practical short-term solutions.The experiences have also shown that, working with partners, the ILO can act swiftly and creatively within its existing mandate, drawing on its dialogue-based approach, neutrality and international labour standards.Notably, the model developed under the Arrangement has already been used elsewhere.However, the improvements made in safety and inspection capacity in Bangladesh could prove ephemeral unless the basic RMG model of deep sourcing and low prices evolves to one of decent work in all its facets (Blasi and Bair, 2019;Kabeer, Huq and Sulaiman, 2019).In particular, experience under the Accord showed the importance of respect for freedom of association in moving towards a fairer wage model and safer work practices.
For its part, the UN treaty process is attempting to turn on its head a legal regime characterized by compartmentalization that abets evasion of responsibility.A change in corporate behaviour with increased respect for human rights is likely to come only from increased corporate accountability, through a variety of means (Bernaz, 2017, p. 297;Simons, 2014).Whether the proposed UN treaty and optional protocol will contribute to this result remains to be seen.At the least, this initiative has already helped to highlight how legal gaps might be narrowed.The proposals need to be legally fine-tuned, while absorbing lessons from the post-Rana Plaza experiments and due diligence experience from a labour and business perspective.If the treaty comes into being, it will become part of a mix of measures still required to prevent workplace disasters and, when this fails, to ensure access to justice and effective remedies for victims.
See, for example, ILO: ILO Centenary Declaration for the Future of Work, International Labour Conference, 108th Session, 2019, Geneva, 2019.Paragraph II.D states that "safe and healthy working conditions are fundamental to decent work".The accompanying Resolution on the ILO Centenary Declaration for the Future of Work requested the ILO Governing Body to consider proposals for including safe and healthy working conditions in the ILO's framework of fundamental principles and rights at work.See also ILO: Resolution concerning decent work in global supply chains, International Labour Conference, 105th Session, 2016, Geneva, 2016.treaty process".A possible new treaty on business and human rights has won the support of hundreds of civil society organizations and deserves examination in the post-Rana Plaza context.What lessons from the experiments in Bangladesh might inform the treaty process?
Center for International for International Legal Cooperation: The Hague rules on business and human rights arbitration, The Hague, 2019.
17 United Nations: UNCITRAL Rules on Transparency in Treaty-based Investor-State Arbitration, United Nations Commission on International Trade Law, Resolution 68/109, New York, 2013.