Voice in Supply Chains: Does the Better Work Program Lead to Improvements in Labor Standards Compliance?

Toward a Broader Understanding of Worker Voice

The notion of unions serving as a form of collective voice is well understood in the industrial relations literature. Direct, management-led forms of worker voice, such as teams and quality circles (Marsden 2013), have been criticized for being used to avoid unions (Watling and Snook 2003), serving as the basis for a “sophisticated nonunion model” (Kochan, Katz, and McKersie 1986), but they have also been increasing in use, alongside a decline in the incidence of indirect collective (union) employee representation (Lavelle, Gunnigle, and McDonnell 2010). Growing attention is being paid to the positive impacts of non-union employee representation (Gollan and Lewin 2013) and the complementary relationship between management-led and other forms of employee voice (Benson 2000).

Prior work examining a wide variety of voice types has been critiqued for focusing strictly on either formal or informal voice mechanisms, when it should instead examine the critical dimensions that may facilitate or inhibit voice (Klaas, Olson-Buchanan, and Ward 2012). Recent research on employee voice is moving toward a broader understanding of the forms that voice can take. Wilkinson, Donaghey, Dundon, and Freeman (2014: 5) defined employee voice as “the ways and means through which employees attempt to have a say and potentially influence organizational affairs relating to issues that affect their work and the interests of managers and owners.” This definition is wide enough to allow for a variety of voice types, including union or non-union, and employer implemented or non-union employee representative systems as a collective form of voice. This is particularly relevant in the context of work in the Global South, where representative forms of collective voice are more difficult for workers to access (Barrientos 2013).

Worker Voice in Global Supply Chains

The fragmented structure of the supply chain can limit opportunity for voice, both unionized collective voice and non-union employee representation (Riisgaard and Hammer 2011). The global market for apparel prioritizes control and flexibility, which makes voice an “inflexible hindrance for competitive advantage” and shifts employer hiring preferences to women, who are considered predisposed to obedience or “voicelessness” (Gunawardana 2014: 455). Low-quality precarious work leaves workers with little power to exercise voice (Josserand and Kaine 2016). Even when mandated by law, the adoption of works councils can be sparse and the quality of social dialogue highly variable, with management able to co-opt the meetings (Wilkinson et al. 2014) and in essence organize workers out of the voice process (Donaghey, Cullinane, Dundon, and Wilkinson 2011).

Repeated calls in the supply chain literature to “bring labor in” (Taylor 2010; Bair and Werner 2015) generally refer to a role for organized labor and union collective voice. Unions face enormous challenges in organizing workers, however, due to outsourcing, the erosion of the direct employment relationship, and downward pressures on firms (Kaine 2014). Workers at the bottom end of the supply chain lack the protection of collective bargaining and may need other types of institutions to enhance their collective voice in enforcement (Tucker 2013).

Emerging forms of transnational institutional arrangements (Marginson 2016) that invoke worker voice as a potential means of remedying the issue of labor standards enforcement have been met with partial success at best. The literature on monitoring CoCs notes a relative absence of unions from monitoring and evaluation, and it documents numerous enforcement challenges (Esbenshade 2004; Locke, Qin, and Brause 2007), as well as a failure to adequately address noncompliance in the area of freedom of association (Locke et al. 2009), an important vehicle for collective worker voice.

The ability of IFAs to develop meaningful co-governance is dependent on a boost in labor power through unions and strong collective bargaining (Donaghey et al. 2011). This dependency is also implicit in other studies that examine, for example, the role of labor agency in improving labor standards along global supply chains (Niforou 2015) and labor control in developing capacity for workplace monitoring (Davies et al. 2011). Although IFAs can help to strengthen union capacity through negotiated relationships with multinational companies, they also rely heavily on union organization and vigilance for implementation, where their roles are only “sketchily defined” (Davies et al. 2011: 132). In addition, IFAs cover only unionized subsidiaries whereas local unions are already affiliated with the global union federation, implying that they may not be protecting those who need it most (Niforou 2012). The adoption and effective regulation of IFAs require strong worker voice mechanisms (Josserand and Kaine 2016), but because of the relative weakness of these mechanisms, IFAs fail to compensate for regulatory gaps (Donaghey and Reinecke 2018), evident in deteriorating labor rights (Levi et al. 2013), freedom of association, and collective bargaining. Their sustainability is in part dependent on greater involvement of all relevant actors (Niforou 2012).

In multi-stakeholder initiatives (MSIs), incentives are at play that could mitigate governance failures associated with other approaches to private regulation. Buyers have a reputational incentive to participate (Oka 2012), manufacturers have the potential for increased orders, and governments see potential for economic gains. Unions also have an incentive, knowing that this forum for social dialogue puts pressure on their employers to be accountable to buyers. Some argue that private regulation cannot be implemented without the involvement or consideration of labor (Kolben 2011; Ruggie 2011). Although MSIs give labor a voice at the table, this role is generally limited to being a “passive object that needs to be taken into consideration, managed and at best consulted” (Riisgaard 2009: 326).

Moreover, MSIs do not always include all stakeholders. The Ethical Trading Initiative is an alliance of companies, trade unions, and NGOs, but primarily controlled by business (Donaghey et al. 2014), leaving out local manufacturers and governments. In his study of the Fair Labor Association, a collaborative effort between universities, civil society organizations, and companies, Anner (2012) found that corporate-influenced MSIs are more likely to focus on violations of minimal standards rather than emphasizing workers’ rights to form democratic and independent unions, explained by their concern with upholding legitimacy and protecting against the risk of consumer campaigns. He further argued that MSIs that are not overly corporate influenced, such as BW, could address this limitation through greater focus on freedom of association, which has much to do with workers’ voice.

Two crucial differences between BW and other MSIs and IFAs are the inclusion of labor at all levels of the program and the inclusion of public actors. Previous research points to the need for union engagement at all levels for effective regulation of corporate social responsibility (CSR) (Harvey et al. 2017), IFAs (Davies et al. 2011; Niforou 2014), and MSIs (Riisgaard 2009), with implications for better enforcement outcomes. BW stands apart by incorporating this practice into its program design, adoption, implementation, and evaluation. Voice is enabled at the global level through the involvement and advisory role of global union federations (GUFs) in program design, in addition to their role in developing technical capacity among and between local unions at the national level. BW’s approach serves an essential function in promoting voice at the global level (Niforou 2014), while at the same time connecting to the local level where labor can play a role in ensuring effective monitoring on the ground (Davies et al. 2011).

Voice is further enabled at the local level through engagement with organized labor as an active stakeholder, addressing a limitation of other MSIs that either leave labor out or include them as “passive objects” (Riisgaard 2009: 326), and moving beyond the “sketchily defined” roles outlined for unions in IFAs (Davies et al. 2011: 132), both of which inhibit labor’s ability to exercise voice. Although BW engages unions in an active role at the global and local level, it does not depend heavily on strong unions and collective bargaining in the same way that IFAs do (Donaghey et al. 2011), but rather plays a facilitative role, providing impetus for weak and/or fragmented unions to come to the table to engage in social dialogue with other actors. At the same time that BW assists with developing the technical capacity of unions, it helps to develop meaningful co-governance without relying on union strength alone. Engaging with workers at the factory level through workplace committees, BW also generates the potential to build workers’ capacity (Alois 2016) and develop relationships between workers and managers on the shop floor, which is crucial for fostering worker voice (Gunawardana 2014; Townsend 2014).

In addition to BW’s facilitative role, recent literature points to the importance of the role of public actors in supporting private initiatives (Locke, Rissing, and Pal 2013; Amengual and Chirot 2016), citing the ILO’s involvement in BW as one example (Marginson 2016), in part because of its credibility as a representative and legitimate global overarching authority (Niforou 2014) and its potential to anchor the initiative to prevent capture by powerful stakeholders (Alois 2018). The growing literature triggered by Rana Plaza further reinforces the importance of the role of public actors, illustrated, for example, in the ILO’s role as independent chair of the Accord Steering Committee (Donaghey and Reinecke 2018) or in Germany’s initiation of the multi-stakeholder Textile Partnership, implementing stricter supply chain governance by shifting policy responses away from the firm to the national collective arena (Schuessler et al. 2019).

Despite growing awareness that effective regulation requires strong worker voice mechanisms (Josserand and Kaine 2016), and several attempts to address this issue, a gap exists in the literature regarding the potential role of non-union forms of collective representation and the importance of ILO-supported initiatives that include labor at all levels of design and implementation. Through the provision of an institutionalized form of collective worker voice, though not necessarily union-based, BW applies a broader understanding of voice that goes beyond other MSIs and IFAs to include whatever “ways and means through which employees attempt to have a say and potentially influence . . . their work” (Wilkinson et al. 2014: 5). These crucial differences position BW ahead of other regulatory initiatives, with better prospects for enforcement outcomes.

A particular strength of the research design is that it tracks developments over time. Several informative case studies have evaluated the impacts of private initiatives on enforcement outcomes, for example, following the policy cycle of two IFAs (Niforou 2012) or observing monitoring efforts of buyer codes of conduct over time (Locke et al. 2009). None to my knowledge, however, have tracked their impacts from the time of program implementation through expiration. In examining the only one of nine BW programs to shut down, this article further contributes by generating additional insights into the factors that contribute to the sustainability of BW, and potentially other MSIs.

A second strength of the research design is the inclusion of worker voice. Assessments of private initiatives have traditionally been conducted through interviews with global union federations, local unions, factory management, NGOs, and other actors who may in theory represent the interests of workers, but this approach raises an important methodological issue in research on the importance of worker voice.

Better Work and Performance Improvement Consultative Committees

BW is a multi-stakeholder initiative that engages buyers, governments, manufacturers, and unions in social dialogue around improving labor standards compliance without negatively affecting competitiveness. BW programs are established as five-year projects, after which they are intended to become self-sustaining. Its pilot project, Better Factories Cambodia, launched in 2001 and eventually led to cooperation between stakeholders, improvements in compliance, and increased trade quotas (Polaski 2006). Building on the project’s success, BW was established and rolled out in an additional eight countries: Jordan (2008), Haiti (2009), Lesotho (2010), Indonesia (2011), Nicaragua (2011), Vietnam (2011), Bangladesh (2014), and Ethiopia (2019).

BW Global is housed at the ILO in Geneva and consists of a global advisory committee, a tripartite project advisory committee in each program country, and local offices to administer its monitoring and evaluation services. BW’s compliance assessment tool focuses broadly on eight compliance clusters related to core labor standards (child labor; forced labor; discrimination; and freedom of association and collective bargaining) and conditions of work set out by national law (health and safety; working time; compensation; and contracts and human resources). The assessment tool is applied by BW staff during unannounced visits, and follow-up activities are focused on remediation and training, rather than on penalties or pulling orders.

BW also establishes workplace committees in participating factories, which comprise an equal number of management and worker representatives (both union and non-union). These committees are known as performance improvement consultative committees (PICCs). PICCs are tasked with the development, implementation, and monitoring of factory improvement plans to address areas of noncompliance identified in factory assessments.

Over the past ten years a growing body of research has examined the impact of BW in its program countries. The research has been largely quantitative, using data primarily from audit reports. On the one hand, evidence shows support for BW, contending that it has led to improvements for hundreds of thousands of workers (Polaski 2009; Rossi 2015), and has had a positive impact on wages and human resource innovation (Robertson 2011), factory survival (Brown, Dehejia, and Robertson 2011), verbal abuse (Rourke 2014), sexual harassment (Lin, Babbitt, and Brown 2014), profit maximization (Brown et al. 2015), factory performance (Brown, Dehejia, and Robertson 2013; Asuyama, Fukunishi, and Robertson 2017), communication and relations between workers and managers (Pike and Godfrey 2015), the public labor inspectorate system (Dupper, Fenwick, and Hardy 2016), compliance within factories supplying to reputation-conscious buyers (Oka 2012), and women’s empowerment at the workplace and household levels (Pike and English 2020).

Some researchers credit improvements to ILO monitoring and the establishment of independent arbitration councils (Shea, Nakayama, and Heymann 2010), increased collaboration between stakeholders (Wetterberg 2011), reinforcement of state regulatory institutions (Amengual and Chirot 2016), facilitation of partnerships and building capacity (Alois 2016), and an emphasis on improving inspection, not on changing the rules themselves (Bernards 2015). Others note the benefits PICCs have for factory owners by helping them understand what workers want (Alois 2016) and the tailored advice they offer factories to develop and implement improvement plans (Miles 2015).

On the other hand, evidence also points to the limited impact that even well-functioning PICCs can have on lowering strike rates (Anner 2017a), as well as the negative impact that the “sourcing squeeze” can have on PICCs, whereby lead firms reduce the prices and production times they allot to their suppliers, undermining efforts by PICCs to address cost-sensitive and overtime violations (Anner 2018: 76). BW has also been critiqued for its voluntary nature, in that it lacks state influence (Anner 2017b) and engenders the same “sorting dynamic” that MNCs do, whereby brands are matched to high-standard suppliers and the rest of the buyers are matched to low-standard suppliers, with the result that the overall level of working conditions remains unchanged (Koenig-Archibugi 2017: 4).

In Lesotho, BW faces particular challenges in terms of the industry’s placement in the global supply chain, its vulnerability due to reliance on trade preferences, a weak and constrained state (Godfrey 2015), fear of chasing away investment (Morris and Staritz 2016), union fragmentation, the fact that BW is not mandatory in Lesotho, and that local stakeholder capacity is further challenged by frequent changes in government. At least three features of BW, however, generate momentum for its potential to lead to improvements in workers’ perceptions of labor standards compliance in Lesotho. First, ILO involvement in BW strengthens its credibility in comparison with other initiatives. Second, the structure of BW and its implementation of worker voice mechanisms at the global and local level have positive implications for the possibility of both union and non-union workers to have meaningful input into the conditions that affect their work. Third, BW’s focus on remediation and training has implications for the role that local actors can play in supporting enforcement by focusing on capacity development and improving workers’ access to effective remedies.

Methods

Data Collection and Analysis

Data for this study were collected through focus group discussions with workers in Lesotho’s clothing industry. Union organizers in the two major industrial cities, Maseru and Maputsoe, recruited workers to participate in the focus groups, which were held at the BW office in Maseru and two separate union offices in Maputsoe. I facilitated the focus groups with the assistance of an interpreter whom I have been working with in Lesotho since 2011. In total over the four waves of data collection between 2011 and 2017, I facilitated 55 focus group discussions with 426 workers. Table 1 provides a breakdown of the number of focus groups facilitated by year, including details on participant demographics.

OPEN IN VIEWER

Table 1.

Focus Group Participant Demographics, 2011–2017

Focus group participant demographics	2011	2013	2015	2017
Focus groups facilitated	17	20	10	8
# Workers	143	141	70	72
% workers in BW factories	38%	77%	93%	69% a
# Firms represented (firms)	41	33	18	26
% factories participating in BW	27%	64%	61%	42% a
Female to male ratio (%)	73:27	77:23	69:31	71:29
Average age	35.62	n/a	33.4	n/a
Average years working at factory	4.09	5.0	5.7	6.6
PICC members	n/a	n/a	33%	7% a
Union members	100%	61%	93%	87%
Shop stewards	n/a	39%	66%	41%

Notes: BW, Better Work; BWL, Better Work Lesotho; PICC, performance improvement consultative committees.

a

Formerly, before BWL shut down.

The same general guidelines for focus groups were used in each wave of data collection, with some modification. In 2011, workers were asked to discuss what they liked and did not like about their work, which developed into a discussion about key areas of noncompliance. Workers raised a number of issues, some of which could later be categorized according to the eight BW compliance clusters, and some of which fell outside this matrix. For example, workers frequently raised concerns about the issue of supervisor relations.

In 2013, similar guidelines were followed, except workers were asked to discuss improvements and persistent problems since BWL launched, as well as the impact of BWL on their home lives. In 2015, the key difference was the addition of a question on what workers felt would happen if BW were to leave Lesotho. In 2017, BWL had been shut down for nearly one year. Workers were asked to describe how things had changed since its departure. They were further prompted to discuss continued improvements and persistent problems, and the reasons explaining these changes.

Each focus group was facilitated using simultaneous translation—an interpreter immediately translated what workers said into English, speaking directly into the recorder. This process made it easy to later transcribe the audio files, which I coded in NVivo. I created coding themes based on the research question of how worker perceptions of labor standards compliance were changing over time—before, during, and after BW intervention (and what might explain this).

Voiced but Unheard: Barriers to Voice in Lesotho’s Clothing Industry

Although options exist for workers in Lesotho to exercise voice through both formal and informal as well as collective and individual channels, several barriers to accessing these channels also exist. In addition to institutional and regulatory challenges, cultural and power dynamics that effectively block compliance with labor standards underlie issues of supervision.

Lesotho’s clothing industry is subject to multiple public and private forms of governance, including a national labor code, international labor standards, buyer CoCs, and trade unions. The Lesotho Labor Code, 1992, establishes a legal framework for the protection of labor standards that enables Lesotho to meet its obligations as a member state of the ILO. Although comprehensive in scope, the code is ill equipped to deal with modern developments and is currently engaged in a (slow) process of revision, with technical assistance from the ILO (Godfrey et al. 2019). Further pressure is added to workers’ access to voice channels through enormous backlogs in the Directorate for Dispute Prevention and Resolution, union fragmentation, and associated weak bargaining power.

In addition to these barriers, cultural and power dynamics in the factories have made it difficult for workers to resolve issues with superiors. Workers believed that Basotho (citizens of Lesotho [plural]) were being promoted to HR because of their ability to be harsh with workers. In the Taiwanese-owned factories, two supervisors were appointed per line: a Taiwanese or Chinese supervisor to advise on the technical aspects of the work, and a Masotho (a citizen of Lesotho [singular]) supervisor to “motivate workers.” The latter translated into Basotho workers being promoted based on their ability to be tough on workers, push them, and shout at them in their local language to make sure they met their targets. In South African–owned factories, only Basotho workers supervised the lines. Although workers did not perceive the same role differentiation in terms of Basotho supervisors being used to push workers, they complained that they were rude to them, and set themselves apart, siding with management.

Workers felt they could not approach their supervisors to resolve their issues, in part because many of their complaints were tied to issues with their supervisors, and in part because their supervisors would simply send them to talk to HR. Once there, HR would either side with the supervisors or tell the workers to sort out the issue with their supervisors. This circular pattern of attempting to resolve grievances did not provide any opportunity for meaningful input or access to effective remedies.

“Contracts and HR” is one of the compliance clusters that BW evaluates, and it was an issue frequently raised in focus groups with workers. Workers believed that HR was not only failing as an effective channel through which to resolve grievances, but was also failing to effectively handle the issue of “fake contracts written in English”—fake in that workers could be fired anytime, reflecting fear around job insecurity and a major source of power imbalance in the employment relationship.

In 2011, workers painted a picture of compliance that did not appear to have changed much compared to research on the sector dating back to the 1980s (Pickles and Woods 1989), 1990s (Baylies and Wright 1993), and 2000s (Seidman 2009). The issue of occupational safety and health (OSH) was the top concern for workers, reporting a lack of personal protective equipment (PPE), exposure to harsh chemicals, extreme temperatures, dust inhalation, unsanitary toilets, and locked doors. One worker reported:

Once a man drank the chemicals thinking it was water. It totally burnt his insides. He just stayed the whole night and then left in the morning when there was transport.

Others reported having severed fingers or partially lost hearing, developing chest colds, being attacked walking home in the dark, or knowing workers who died from injuries during a night shift because the company only provided transport during the day. Further examination of the data on OSH issues revealed that “buyers” was one of the most frequently used words. Whenever workers received a face mask or protective gloves, for example, it was because the buyers were coming, after which PPE and other safety provisions were withdrawn. One worker said that “Levi’s is like a semi-God to the Chinese” and described how managers evaded audits by closing down entire sections of the factory:

When the buyers are coming, workers in that section are pulled out and the door is locked so that the buyer will pass without noticing.

Although OSH was discussed as the most severe problem, many workers’ main concern was with their supervisors. The issue of supervisor relations came up in virtually all of the focus groups in 2011. It was, however, much more than an issue of bullying or being rude to workers. Closer examination revealed that, when workers spoke about issues with supervisors, they were ultimately talking about interference in their access to labor standards.

For example, with reference to OSH, some workers were not given PPE regularly because their supervisors “didn’t like them.” Freedom of association was also impeded, with several workers remarking that those who were “strong at the union” would be promoted to be supervisors so that they could “be on their side.” Compensation was negatively impacted by favoritism, with workers perceiving that supervisors would be the ones to tell the manager who should get a raise, and that it was “usually because they like them.” In addition, discrimination existed toward workers based on their health status (Lesotho’s clothing industry has a 40% incidence of HIV/AIDS), with workers stating, “If the supervisor learns from the doctor that there’s an issue with you, if they know you suffer from this and this, they’ll talk about it all the time.”

In summary, workers’ perceptions of compliance in 2011 revealed not only current key issues but also the barriers to voice and underlying root causes of noncompliance. Supervisor relations underpinned many of the issues raised by workers, creating barriers to voice and inhibiting access to labor standards. Lacking opportunity for meaningful input or access to effective remedies, workers were caught in a loop.

Breaking Barriers to Voice: The Importance of Better Work PICCs

The baseline report on BWL in its first year drew heavily on workers’ feedback from the 2011 wave of data collection. The findings of this report directly informed BWL training activities in the subsequent year, which mainly focused on “soft skills” for supervisors and HR managers, as well as OSH and emergency preparedness. By 2013, workers reported that conditions had been improving, with feedback largely focused on improvements in communication and OSH.

Analysis of the focus group transcripts from 2013 revealed that approximately one-third of workers’ feedback on “improvements since BWL” could be accounted for within the broad category of “communication and relations.” On this theme, workers alluded frequently to the PICCs established by BWL in their factories, and how these were improving their lives both at work and at home. PICCs serve as a forum in which workers—in dialogue with other workers and management representatives—can have meaningful input and access to effective remedies. Workers described improvements in communication with their supervisors, which in turn improved their relationship and facilitated effective resolution of their issues. One worker described how the PICC was opening channels of communication in the factory:

The good thing about PICC is that employees are free to lodge their complaints. . . . They would be making sure that everything is fine in the factory. Normally we take complaints to management, then I take back the answers to workers. . . . Management is willing to work hand-in-hand with PICC.

Another worker described how this improvement in communication through the PICC was beginning to break down barriers to labor standards compliance.

Since BW, every little issue, we manage to solve it. That why I don’t really have many problems anymore. . . . In PICC meetings, when we request a meeting with the manager, we certainly get an appointment.

Other workers reported many improvements being made since BWL “because we have the PICC,” for example, more frequent provision of PPE, improved cooling systems in the summer, immediate attention to slippery floors, advance notice for working overtime, receiving wages on time, and being allowed to leave at noon on payday (for safety reasons). Through the PICCs, workers believed they were able to address their workplace concerns, effectively breaking down the barriers to labor standards compliance that had been created by poor supervisor relations. The following excerpt from a focus group in 2013 is illustrative of the connection between improvements in communication and the resulting improvements in other compliance items:

In my factory . . . the toilets were dirty . . . we were forced to stay for 6 p.m. overtime and sick leave was not paid. As a shop steward, I was not free to approach management about any issues arising from workers. Now everything is going accordingly because I am now able to approach my employer and discuss relevant issues . . . because I am part of the PICC.

In addition to improvements in communication, approximately half of workers’ feedback on “improvements since BWL” was related to improvements in OSH. In some cases, workers were receiving face masks regularly, pregnant women were given mats to stand on, people with illnesses were being given a tea break to take their medication, some heaters had been installed in the factories, and many workers were given hot water at 10 a.m. A number of workers had received firefighting training and, in some factories, aisles were cleared of clutter, doors unlocked, and workers briefed on escape plans.

One of the unique findings from the 2013 data was the degree to which workers perceived that BWL was impacting their home lives. Several workers described how they were sharing responsibilities at home and resolving conflict with partners rather than bottling it inside, which they previously had a problem with before BW, going home stressed and “taking it out on everyone at home.” One woman remarked,

With my little understanding I have about BW, I know that we as women, we know a lot of things. Therefore I pass that on to my husband and tell him about how things are supposed to be done . . . in order to build a proper family.

Others reported that they were budgeting their wages better and keeping their communities clean and safe, applying whatever they could from BWL trainings. One worker referred to a PICC training on nutrition that gave her skills to look after sick people, including knowing types of food to give them to have strength.

In summary, workers’ perceptions of labor standards compliance in 2013 revealed that the major improvements were in the areas of OSH and communication and relations, related to the establishment of PICCs. Through the PICCs, when workers perceived noncompliance issues they could immediately voice their concerns and have them resolved. The PICCs were also beginning to break down the barriers to labor standards created by poor supervisor relations, thereby establishing channels for workers to have meaningful input and access to effective remedies.

To triangulate these findings, I obtained copies of the three compliance synthesis reports produced by BWL during the life span of the program. Assessments are carried out by BWL and based on a comprehensive set of 250 questions that cover core labor standards and national labor law requirements. Each assessment is based on observation, document review, and interviews with management and workers.

The first and third assessment periods (beginning March 2011 and January 2013) roughly align with the period of time in which the first (May 2011) and second (September 2013) waves of data were collected for this study. The second assessment period fell somewhere in the middle (beginning January 2012). Similar to my findings in 2011, the first compliance report indicated significant need for improvements in OSH, as well as contracts and HR. The report noted that the work of the PICCs was still at an early stage in which a better understanding of the role of the PICC needed to be generated and dialogue structures between management and workers needed to be strengthened (Better Work Lesotho 2012).

The second report showed some improvements but problems persisted in most areas, including OSH, which remained high on noncompliance (Better Work Lesotho 2013). This finding aligns with the focus group data, which pointed to a number of compliance issues—in particular in the area of OSH—in the first year or two of the program. Given that it was still the early stages of BWL, the finding that compliance issues were still prevalent is not entirely surprising.

Similar to my findings in 2013, the third report showed that 11 out of 13 factories that had been assessed more than once had improved their noncompliance rate, in particular in the area of OSH. The report credited advisory visits and worker–management discussions during PICC meetings for helping to bring about improvements in this area (Better Work Lesotho 2014). Appendix Figures A.1 and A.2 illustrate the changes highlighted in the second and third compliance assessment reports. Unfortunately, no additional reports were published, which would have served as a valuable check on the findings presented here on the role of PICCs in facilitating compliance over time.

Keeping the Barriers Down: The Limitations of Better Work PICCs

Whereas workers experienced a number of improvements in labor standards in the first few years of BWL, things seemed to change by the time of the third wave of data collection in 2015. Similar to the findings in 2013, workers’ focus group feedback revealed that the biggest improvement in 2015 was in the area of OSH—for example, being provided with PPE more regularly, having safer work environments, and feeling cared for by being provided with heaters, hot water, and in some cases smoking breaks. It was also the most persistent problem that remained, however. Some workers reported that their employers were starting to repeat old patterns of providing PPE “for buyer visits only.” Although BWL conducts unannounced audits only, one worker described how “it’s a hearsay thing” and that someone from one factory will simply WhatsApp the managers at another factory, saying, “the BW people are here so you better start cleaning.”

Other issues were contributing to a return to old patterns, including the challenge of supervisor turnover on the long-term impacts of their training, in particular with regard to how supervisors treat and communicate with local (Basotho) workers. In 2013, workers spoke about the positive impacts of training and had begun to notice some improvements in supervisor relations. By 2015, the tide had shifted somewhat, with one worker reporting that “they treat the workers like they used to treat them before BW arrived.” For example:

In the olden days, there were reports on sexual harassment and discrimination. The only problem is that most of the trained Chinese supervisors, when their contracts expire, they normally leave the country and then they hire the new ones. Those new ones are untrained and it’s still a problem.

The frequency with which workers spoke about PICCs also decreased overall. In some discussions, workers reported that their managers were beginning to interfere somewhat with the PICCs, for example, listening to input from non-union workers only, or selecting workers who they did not think would pass on the information to other workers. As one worker described:

They would choose a non-union worker, someone who was very dumb, who would not be able to address the workers about what they have learned from the meeting.

At the time this data was collected (December 2015), six months remained before the official closure of BWL (June 2016). Its termination had already been announced and BW was making efforts to develop a sustainability strategy and transition plan alongside the country program–level advisory committee (O’Brien 2015). Not all workers were aware that this was coming down the line, and those who were aware expressed deep concern for the future of their working conditions after BWL’s departure.

Indeed, conditions continued to worsen after BWL shut down. In focus groups in May 2017, most workers reported that their PICCs were either nonexistent or dysfunctional. In one focus group, workers from five different factories said that their PICCs were useless, made up of either office people, supervisors, or union members working closely with the employer—all of whom workers felt sided with management and did not represent their interests. One of the workers who was also a PICC member said that her PICC was “still operating but declining” due to a communication breakdown between management and workers. She described the impact of BWL’s departure on working conditions in the context of a weakening PICC:

It’s dragging their feet . . . management does not attend the meetings, like before while BW was still around. During BW, the management were taking care of the workers, playing a major role towards their needs, but since then it’s changed. . . . Within PICC, if [workers] bring their grievances, [management] will just tell them they’ll take action but there is no action. It’s happening more often that you get fired without explanation, or you have a hearing and there is no reason for that. Even the production manager can just pick a person and tell that person that I do not like you, I won’t work with you. It was not like that when BW was there.

This re-emergence of barriers to voice was linked with a decrease in compliance. Although some workers still received PPE, they were receiving it much less frequently than they were during BWL, for example, receiving a face mask every four months instead of every month. Some reported that other improvements had ceased altogether: first aid kits were no longer filled, exit doors were locked, fire drills ignored, and so on. One worker described the impact of BWL’s departure:

Before BW came, things were pretty bad. BW came, it fixed everything, things were good, we were happy. Now that BW has left, we are going back to square one.

Although some workers said that the improvements in OSH were continuing after BWL, they emphasized that issues with communication and relations were creating bigger problems for them. One worker described this in the context of communication with supervisors:

During BW, supervisors and managers were given training on communication skills. They were able to communicate with the rest of the workers. There was no noise. It was a skill that they accumulated. But now it’s not the same. It deteriorated. There’s no communication between the workers. It’s terrible . . . to the point that we don’t feel like working.

In this case, the breakdown is attributable to a lack of training for supervisors in the wake of BWL’s departure. These types of breakdowns, however, were also present during BWL. As the 2015 findings indicate, supervisor turnover and a lack of ongoing training for supervisors is a major challenge for long-term sustainable improvement in this area.

The themes of punishment and revenge came up in a couple of the focus groups, as part of a discussion on why conditions were worsening. This finding is potentially reflective of management’s resentment of the pressure BWL was able to put on the employment relationship, more traditionally typified in this context by power distance, managerial control, worker obedience, and silence. One worker reported that her manager told her, “Your BW is no longer here, so we are not going to do anything.” Another described how this issue was affecting workers in practice:

It’s like we’re being punished or treated as slaves because management does not treat us in the right way. While BW was still around, they would treat us nicely, talk to us in a polite way. . . . But now they don’t care. They just talk to us as they please.

Beyond the workplace, workers also perceived that the departure of BWL was having a negative impact on their home lives. Several workers said they felt comfortable when BWL was there, that they woke up looking forward to going to work, knowing that management would be abiding by the rules. One worker said that now, after BWL, it is “difficult to even wake up”; another, that “it’s a trouble for me to even wash myself to go to work.” One worker described the impact on his life at home:

We are treated like slaves again, and it’s affecting our families. When BW was still around, we felt free, we were happy that, even when we wake up in the morning, we feel like, okay, I’m waking up, I’m going to work now. There was peace in the house. But now, even when you’re supposed to wake up in the morning, it’s a struggle because in your mind you’re saying like I’m going to hell. And we fight with our partners on a daily basis, in the morning and in the evening, because of the stress that we encounter while we’re at work. We’re no longer free at all.

The departure of BWL led to an overall decline in workers’ perceptions of compliance, in large part attributable to the deterioration of the PICCs and other negative impacts on communication and relations. Although evidence exists that some workers continued to benefit from OSH improvements that were made during BWL, most workers were receiving little to no PPE and noticed a return to unsafe work environments. Workers no longer felt as free, comfortable, or happy to go to their jobs, as they did during BWL, and noticed this was generating more conflict at home.

Discussion and Conclusions

BW sets itself apart from other initiatives through its inclusion of labor at all levels, including through PICCs at the local level. Previous research has demonstrated the potential role of PICCs in improving labor standards by helping managers understand what their workers want (Alois 2016), providing tailored advice to factories on the development and implementation of improvement plans (Miles 2015), and serving as a democratic forum for worker representation, protection, and empowerment (Anner 2017a). The findings of this study add to previous work on the role of PICCs by moving the analysis to the shop floor and utilizing worker feedback to understand the process through which PICCs can lead to improvements in workers’ perceptions of compliance. Specifically, PICCs serve to break down barriers in the contested space of supervisor–worker shop floor relations, opening up channels of communication that facilitate improved labor standards. The findings further extend previous work by demonstrating how communication and capacity-building within the PICCs lead to improvements beyond the factory, in workers’ home lives.

Previous studies of other collective mechanisms for regulating labor standards in global supply chains have identified the importance of, and challenges to, incorporating voice in labor standards enforcement, for example, Harvey et al. (2017) on CSR and Williams, Davies, and Chinguno (2015) on IFAs. This study underlines the importance of voice, but differs from the somewhat normative focus on the role of unions in prior research. Rather, this study applies a broad definition of worker voice to include both union and non-union forms of collective representation, drawing on through-time data to examine closely the ways and means by which BW and PICCs provide opportunities and tools for workers to have a say in the decisions that affect their work.

Although the empirical analysis highlights the importance of PICCs as a joint mechanism for enforcing labor standards in global supply chains, it also reinforces concerns about the sustainability of BW (Oka 2010; Bair and Gereffi 2014). The presence of the ILO as a type of standing reference group in BW implementation and monitoring helped to establish a continuous working program, treating its mandate as “enforceable objectives, rather than as aspirations” (Davies et al. 2011: 136). When BWL shut down, however, conditions started to worsen. This occurred despite efforts by BW to develop a sustainability strategy and transition plan alongside the country-level advisory committee (O’Brien 2015). Future research should focus on understanding the role of voice in structuring and implementing a transition plan for BW. Not only would this contribute to a gap in the literature on the factors and processes facilitating sustainability of transnational arrangements but it would also equip policy actors at BW to design strategies for achieving one of the major goals of their mandate: to operate in a country for five years and in that time develop local capacity sufficiently to hand over the reins.

The findings indicate that voice declined when it became less representative. A cross-cutting theme of recent research on the role of voice in supply chains is the focus on the democratic nature of the initiatives that purport to espouse worker voice. Harvey et al. (2017) considered the democratic nature of trade unionism, and the legitimacy of trade unions to be involved in deliberation over CSR. Niforou (2014) borrowed four core democratic principles—representation, legitimacy, accountability, and transparency—and discussed them in relation to the ability of IFAs to address the global governance deficit. Anner (2017a) similarly drew upon democratic principles—elect, represent, protect, and empower—to develop criteria for well-functioning worker committees. At its core, this perspective implies that workers should be able to have meaningful input into the decisions that affect their work, be free from retaliation, and be adequately represented. It does not create a new actor (Hammer 2005) or new regulations, but rather represents a type of “conversion” whereby established actors (in this case, workers) draw on capacities within institutional arrangements that have hitherto lain dormant, to alter their pre-existing trajectory (Marginson 2016).

This study has one main limitation: Incomplete data did not allow me to analyze associations between mechanisms of appointment/election to the PICCs and perceptions of their effectiveness. Future research on PICCs in other BW countries could collect and systematically analyze data on the effectiveness of PICCs across a number of labor standards outcomes, applying Anner’s (2017a) criteria for well-functioning PICCs. Access to data on PICC formation date, number of PICC meetings held, and content analysis of meeting notes would further strengthen analysis of the role that PICCs play in enabling worker voice, with prospects for better labor standards enforcement outcomes down the supply chain.

This study addresses a call for future research to assess the local impact of transnational labor arrangements, such as the ILO’s BW, in the Global South (Niforou 2014). In addition to being inclusive of labor at all levels, BW sets itself apart from other initiatives through its inclusion of public actors. The findings demonstrate an important role for the ILO in supporting private initiatives, although it raises questions about the obstacles preventing a transition away from dependence on the ILO toward a more collaborative and sustainable approach to labor standards enforcement in supply chains. In order to secure sustainability post–ILO withdrawal, the findings point to the need to ensure that local and national public actors are equipped to step up. If we are to understand not just outcomes for labor standards but the processes through which workers perceive changes, further attempts to advance scholarly insights into the role of voice in supply chains must apply a broader understanding of worker voice.