APPAREL AND FOOTWEARON A SATURDAY AFTERNOON IN MARCH 1911 A FIRE BROKE OUT ON the eighth floor of a garment factory in New York City. The blouses and scrap materials at the Triangle Shirtwaist Factory burned quickly. Workers struggled to get out of the building, blocked by locked doors, flames, and the collapse of the fire escape, and many resorted to jumping. In total 146 people, mostly young women, died. The Triangle fire became the deadliest accident in the American garment industry and inspired a generation of activists, unionists, and reformers.
Eighty-two years later, in 1993, the Kader toy factory in Thailand experienced a similar fire, killing 188 workers, again mostly women. This time activists raised questions about the responsibilities of retailers and brands such as Toys R Us and Hasbro that were sourcing from Kader. In the ensuing years, as the apparel and footwear industry aggressively embraced the supply chain revolution, similar questions began to be asked about the responsibilities of Nike, Walmart, and others. The rise of anti-sweatshop movements in the United States and Europe led most apparel and footwear brands to adopt “codes of conduct” and to promise to better police their suppliers, spawning a growing field of “corporate social responsibility” (CSR).
Mounting questions about these activities’ effectiveness became especially acute in 2010–2012, when a series of factory fires in Bangladesh killed nearly two hundred people who had been producing apparel for H&M, Tommy Hilfiger, C&A, Walmart, and others. Then, in Bangladesh in the spring of 2013, the Rana Plaza complex of factories collapsed, killing more than eleven hundred workers—once again, mostly women. This was the worst industrial accident in the history of the garment industry, and it happened despite the codes of conduct and CSR programs of Primark, Benetton, the Children’s Place, and other companies that were sourcing there. In the same industrial district, a factory producing sweaters for H&M had collapsed eight years earlier (Miller 2013), but the scrutiny that ensued was clearly not enough.
In this chapter we examine the emergence and influence of voluntary labor standards (codes of conduct, factory auditing, and certification) for apparel and footwear production. As these industrial accidents suggest, despite the rise of CSR and many calls for “ethical fashion,” clothing sold in the United States and Europe comes with few meaningful guarantees of decent production. The reason, we argue, stems from a combination of factors—namely, the power of companies in standard-setting initiatives, questionable factory auditing, weak links between companies’ compliance and sourcing decisions, and a highly mobile industry that has relentlessly chased low prices at the expense of decent labor conditions. These forms of “unruliness” have greatly limited improvements and undermined certification of “good” factories. Some firms have done little to make their codes of conduct meaningful, but even the brands and retailers that have become known as leaders in the field of CSR, such as Nike, the Gap, and H&M, have often worked at cross-purposes. They have asked suppliers to improve but have continued to push for low prices and quick delivery times and to move in search of them. Promoting “fair” labor in the apparel and footwear industry has mostly meant slapping standards onto existing, highly unequal, structures of production.
Overall, this chapter shows that codes of conduct have regulated the conduct of suppliers, although only marginally and unreliably. But they have not regulated the conduct of brands and retailers themselves, and this is their greatest weakness. We begin by describing the structure of production and consumption, which has fostered the rise of global rule making but also limited its impact. Then, examining the formation of monitoring and certification programs, we show how a range of different programs, serving different constituencies, sought to define rules for “fair” labor. We then examine some marginal impacts of these rules and explain their larger failings by showing how implementation and industry dynamics have remained “unruly.” A closer look at factories in China further illustrates how local contexts can generate serious challenges for global rule making. Finally, given these failings, we consider whether some emerging alternative models might provide more meaningful choices for conscientious consumers.
American and European (EU-27) households spend an average of $1,700–$1,800 per year on apparel and footwear (US Department of Labor 2013; Eurostat 2009, 182). (In contrast, even with rising consumer expenditures in China, urban households there spend an average of around $260 per year on apparel and footwear [Li & Fung 2012].) Approximately half of the world’s exported apparel ends up in the United States or EU countries (and more than this if one counts cross-border trade within the EU).1 Since the United States and some European countries require the country of origin to be shown on apparel and footwear labels, consumers have been able to wonder, as journalist Kelsey Timmerman (2012) puts it, “Where am I wearing?” The answers to that question have shifted over time. To understand why the global apparel and footwear industry have been mobile, and why they have generated so many concerns about sweatshops and labor rights, a brief sketch of its structure and operation is necessary.
A tendency to “sweat” the workforce—that is, to extract as much as possible through long hours and low pay—is to some degree built into the structure of apparel production. In industries where employers have large fixed costs—for heavy machinery, for instance—they are generally motivated to invest in workers’ productivity. They have already sunk a great deal of money into the machinery, so they need to get as much production from it as possible. As Michael Piore (1997) points out, since apparel production is a labor-intensive activity that cannot easily be mechanized, employers have little reason to maximize hourly productivity, especially if workers are paid by the piece, as they still are in many parts of the garment industry. Employers can simply push workers to put in longer hours when high output is necessary and fewer hours during low seasons. Seasonal swings and rapidly changing fashions can further exacerbate instability, resulting in long, sometimes unpaid, overtime and high-stress working conditions (Anner, Bair, and Blasi 2013; Bonacich and Appelbaum 2000). Footwear production is slightly more mechanized, and textile production (that is, the manufacturing of fabrics) is significantly more so, which means capital investments are greater for these types of firms. Still, the apparel, textile, and footwear sector as a whole is among the most labor-intensive of all industries. Labor costs contribute only 2 to 6 percent of the final retail price of apparel (Collins 2003; Nova and Kline 2014), but employers still compete fiercely over these small differences.
Contracting relationships, rather than vertically integrated production, have long been central to apparel production. Brands and retailers that are familiar to consumers typically do not own the factories where production occurs, instead relying on independent contractors and suppliers, who may themselves subcontract some parts of the job. This feature of the industry is not new. Even when it was primarily nationally or locally based, apparel production was organized through contracting relationships (Bonacich and Appelbaum 2000; Uzzi 1996; Waldinger 1986). But globalization amplified this tendency, as the “branded manufacturers” that did own some of their own factories (e.g., Levi Strauss, Phillips Van Heusen, VF) increasingly shut them down to join the “supply chain revolution” of the 1990s.
Apparel and footwear production is often seen as the prototypical “buyer-driven commodity chain,” meaning firms that are closest to consumers (retailers and brands) are the powerful “lead firms,” with the power to shape styles, technologies, and prices throughout the supply chain (Gereffi 1994). These lead firms include general retailers that sell “private label” apparel (e.g., Walmart, Carrefour, Macy’s), specialty apparel retailers (e.g., the Gap, H&M, Zara), “branded manufacturers” (e.g., Levi Strauss, Phillips Van Heusen), and “branded marketers” who focus almost exclusively on design and marketing (e.g., Nike, Liz Claiborne). Growing concentration in the retail sector has put department stores and discount retailers in an especially powerful position (Collins 2003), with the largest discount retailers accounting for most apparel sales in the United States (Appelbaum and Gereffi 1994). In the athletic footwear industry the power of brands such as Nike, Adidas, Puma, and New Balance has not yet been rivaled by footwear retailers.
If suppliers are “captive” to these powerful buyers, one might expect the buyers to be able to force their suppliers to comply with labor standards. As Richard Locke and his colleagues (2009) point out, advocates for fair labor standards have often assumed, incorrectly, that this type of power was likely to bring about factory-level compliance. As some suppliers have grown larger and taken on more complex manufacturing tasks, scholars have questioned whether lead firms are fully in the driver’s seat (Gereffi, Humphrey, and Sturgeon 2005) and whether collaborative relationships might have more potential for improving labor conditions (Frenkel and Scott 2002; Locke, Amengual, and Mangla 2009). In footwear manufacturing, Asian transnational manufacturing companies (e.g., Ching Luh, Pou Chen/Yue Yuen) have become larger and more powerful, and some apparel manufacturing companies (e.g., Luen Thai and the massive sourcing firm Li & Fung) are following suit (Appelbaum 2008).
Nevertheless, the degree of inequality between brands and retailers on one hand and the factories that supply them on the other remains stark, especially in the apparel industry. Retailers and brands continue to capture an outsized share of the profits from global production, and they are often able to nimbly shift their orders between suppliers and locations to drive down prices. This becomes clear when looking at the prices that manufacturers receive. Even when factories have “upgraded” their manufacturing capacities, the prices they can command has often fallen (Schrank 2004). Based on calculations made by Mark Anner and his colleagues (2013), the real unit price of men’s and boy’s cotton trousers from the top four exporting countries (China, Mexico, Honduras, and Bangladesh) declined between 2000 and 2010, and the unit price of all apparel imported into the United States fell by almost half between 1989 and 2010. In addition, the rise of “fast fashion” discount retailers, such as H&M and Forever 21, has meant that factories are under pressure not just to sell at low prices but also to adapt to frequent changes in orders and short turnaround times. These pressures have left little room for significant improvement in labor conditions.
When one says “workers” in the apparel and footwear industry, more often than not this means women, especially young women. Young women have long been the preferred workers in labor-intensive manufacturing industries (e.g., apparel and electronics), in large part because managers expect them to be docile and willing to work for low wages (Cowie 1999). As apparel and footwear production globalized, it gravitated especially to places where patriarchy was strong—first South Korea, then Indonesia—where young women from rural areas were expected to be “good daughters,” saving for their dowries and sending money home to their families (Enloe 2000). But beyond simply taking advantage of gender inequality, companies have often fostered gendered images of workers. As Jane Collins (2003) puts it, “Managers in the apparel industry have historically relied on gendered ideologies of sewing work to devalue women’s skill and lower their wages” (16). Leslie Salzinger (2003) goes further, showing how shop floor relations in Mexican maquiladoras produce particular notions of femininity. When production demands have outstripped the supply of women who are willing to work for low wages, apparel firms have proven willing to employ men (Ross 2004), but apparel production has continued to be viewed—and paid—as low-wage, feminized work.
Government policy has also played an important role in the globalization of apparel and footwear production. As Ellen Israel Rosen (2002) has shown, US military and foreign policy interests during the Cold War shaped the first wave of globalization, taking apparel (and later footwear) production to Asia—first to Japan, then to Hong Kong, Taiwan, and South Korea—as a buffer against left-wing parties and the specter of communism. In the 1980s, US trade policy began to strongly promote apparel production in Central America and the Caribbean. Many manufacturers and specialty retailers shifted from domestic production to this region, taking advantage of wages that were less than fifty cents per hour in Guatemala and El Salvador, sixty to seventy-five cents per hour in Honduras and the Dominican Republic to eighty-eight cents per hour in Mexico at the beginning of the 1990s (Bonacich and Waller 1994). For European firms, production in Eastern Europe and North Africa played a similar role (Gereffi 1999).
Discount retailers had long had extensive ties to suppliers in Taiwan and Hong Kong, but by the late 1980s they were increasingly sourcing from mainland China as well (Cheng and Gereffi 1994). Korean, Taiwanese, and Hong Kong–based owners of apparel factories were also investing in the Philippines, Indonesia, and Thailand, which soon became key sourcing destinations for global brands (Bonacich and Waller 1994). Apparel industries in Southeast Asia, Central America, and Eastern Europe soon found themselves in heavy competition with factories in China and Vietnam. As of the early 1990s labor costs in China were well below those in most Southeast Asian countries (Yang, Chen, and Monarch 2010), and by 1994 mainland China had become the world’s top apparel exporter.2
Trade policy also kept the industry geographically dispersed, as opposed to gravitating to just a few countries. The Multi-Fiber Arrangement (MFA), a set of trade agreements originating in the 1970s, limited the extent of apparel and textile imports to the United States, Canada, and European countries that could come from any single country. Exporting country governments received negotiated quotas for imports to these markets, which they could then distribute to firms. As a result, even as China became a major apparel exporter, it was limited by its quota allotment. The MFA fostered the growth of garment factories in Lesotho, Mauritius, and other hinterland locations that had enough infrastructure to take advantage of their quota allotments. Ultimately, however, the MFA was judged to be inconsistent with the growing project to promote free trade, under the auspices of the General Agreement on Tariffs and Trade (and soon the World Trade Organization). The MFA began to be phased out in 1995 and the final set of quotas was eliminated in 2005.
In the post-MFA period China’s apparel exports have continued to grow and the small hinterland industries have declined. But the industry did not gravitate entirely to China, in part because wages there began to rise. With rising wages in China and a wave of strikes in Vietnam, apparel production in places such as Indonesia and Honduras continued to be competitive. But brands and retailers have increasingly looked to South Asia as the next key location. Low tariff treatment for the European market made Bangladesh especially attractive for European firms, helping the value of Bangladesh’s apparel exports to increase by 294 percent from 2000 to 2011 and its share of total world apparel exports to increase by a whopping 713 percent between 1990 and 2011. Less dramatically, India’s share increased by 49 percent and Pakistan’s by 17 percent from 1990 to 2011.3
In contrast, Thailand’s share declined by 42 percent from 1990 to 2011. In Mexico and Indonesia apparel exports grew in the 1990s but then declined—sharply in Mexico (by 74 percent from 2000 to 2011) and more modestly in Indonesia (by 18 percent from 2000 to 2011). China’s apparel exports continued to grow, and Indonesia remained competitive because wages were rising much more slowly there than in the coastal regions of China. But monthly minimum wages of $120–$140 in the main manufacturing regions of China and Indonesia in 2010 were far higher than in Bangladesh, where the monthly minimum wage stood at $43, even after a substantial increase in response to workers’ protests.4
Wages are not the only factor that makes a particular country attractive or unattractive to global apparel brands. Brands also consider factors such as labor productivity, whether suppliers can deliver the appropriate quality and quantity quickly and reliably, and how tariff rates (which may vary by country and product) will shape their final costs. Governments in developing countries have sought to attract international orders by building export-processing zones (EPZs) with improved infrastructure, favorable tax treatment, and sometimes exemptions from national law. Still, enough governments have been competing in this way that labor costs have often been decisive in shifting apparel production from one frontier to another.
This networked, stratified, global, and quite mobile system of production has hovered over, and often haunted, calls for “corporate social responsibility” and “ethical fashion.” Large parts of the global apparel and footwear industry are geared toward serving American and European consumers, and as described in chapter 2, some nontrivial portion of these consumers does care about labor conditions. But this does not mean that conscientious consumerism has transformed apparel and footwear production, as the rest of this chapter will show.
Inequalities and exploitation in the global apparel industry might have remained largely hidden were it not for the rise of a new anti-sweatshop movement. The Triangle Shirtwaist Fire provoked an earlier wave of anti-sweatshop activism, union organizing, and state-building in the United States (see Stein 1977). Yet this mid-twentieth-century “social compact” began to break down as US-based apparel companies found new opportunities to move abroad (Esbenshade 2004). With apparel production becoming more globalized, and work in domestic garment districts becoming more precarious, activists from unions and from religious, immigrant rights, and human rights organizations began building a new anti-sweatshop movement in the 1990s. They focused to some degree on US government trade policy, which had facilitated the migration of the industry (e.g., Krupat 1997), but they quickly seized on the strategy of “naming and shaming” well-known brands and retailers.
In early campaigns, immigrant rights activists named and shamed fashion designers such as Jessica McClintock for using domestic sweatshops (Louie 2001). UNITE (the Union of Needletrades, Industrial, and Textile Employees) and the National Labor Committee (NLC) were central to a number of campaigns that followed. Starting in 1995 the two cooperated on a campaign targeting the Gap, among others (JC Penney, Eddie Bauer, and Target), over the repression of union organizing at the Mandarin apparel factory in El Salvador. Walmart and Nike soon became the top recipients of the sweatshop stigma, mostly due to exposés of child labor, harsh treatment, and poverty-level wages in Asia. On college campuses in the United States, students affiliated with United Students Against Sweatshops rallied against collegiate licensing deals that made universities complicit in labor rights abuses. In Europe the federated Clean Clothes Campaign used a mixed “oppose and propose” strategy to press firms such as H&M, Adidas, Karstadt-Quelle, and Carrefour to take responsibility for labor conditions in their supply chains (Ascoly, Oldenziel, and Zeldenrust 2001; Bair and Palpacuer 2012). With the exception of a boycott against child labor in Bangladesh, activists rarely called for outright boycotts, but they did seek to demonstrate that consumers, investors, and business partners would not remain silent. As the NLC’s Charles Kernaghan said of the Gap campaign, “[We are] not calling a boycott, but we want the Gap to get the message” (qtd. in Tosh 1995). In total, in the United States more than 25 percent of large lead firms in the apparel and footwear industry were targeted in anti-sweatshop campaigns between 1993 and 2000, and many faced persistent pressure (Bartley and Child 2014).
Anti-sweatshop campaigns catalyzed an array of rule-making projects as companies, NGOs, and unions vied to define the responsibilities of lead firms and the meaning of fair labor conditions. Initially, when faced with sweatshop accusations, brands and retailers minimized their responsibilities, emphasizing that they did not directly own the factories or that apparently shocking stories were not what they seemed. Walmart’s CEO infamously replied, “The pictures you showed mean nothing to me,” when confronted with images of child laborers in a Bangladeshi factory (qtd. in Ramey and Barrett 1996, 10). But as anti-sweatshop pressure mounted, companies began to accept at least a limited responsibility for improving conditions in their supply chains. After years of such pressure, Nike’s CEO admitted that the company “has become synonymous with slave wages, forced overtime and arbitrary abuse” (qtd. in Cushman 1998). Brands were beginning to accept an ethical responsibility for their supply chains, but they have fiercely resisted the idea that they might have a legal responsibility for their supply chains (Bonacich and Appelbaum 2000; Shamir 2004).
Levi Strauss, Nike, Reebok, and the Gap were among the first companies to adopt “codes of conduct” for their suppliers, but by the end of the 1990s the vast majority of large lead firms in the apparel and footwear industry had followed. Companies targeted by anti-sweatshop campaigns were hoping, as one industry advisor put it, to “put a muzzle on these watchdog groups” (Rolnick 1997), but even those that had so far largely flown under the radar (e.g., Jones Apparel, Talbots, and VF—maker of Lee, Wrangler, Jansport, and many other brands) quickly adopted codes in hopes of preempting campaigns or protecting their reputations (US Department of Labor 1996).
Codes of conduct (sometimes known as ethical sourcing policies) typically ask suppliers to eliminate child or forced labor, to ensure that working conditions are safe, that workers receive at least the legal minimum wage, and that working hours are kept within a given standard (often sixty hours per week). Many codes also ask suppliers to respect workers’ freedom of association—that is, their right to join unions—and invoke conventions developed by the International Labor Organization (ILO).
At first lead firms adopted these rules without taking serious steps to enforce them. Some asked quality control personnel to also check compliance with labor standards, while others sent poorly trained external auditors to the factories. Nike initially used auditors from Ernst & Young and PricewaterhouseCoopers, who were poorly suited to gather accurate information from young migrant workers or to assess health and safety conditions (O’Rourke 1997, 2002). Labor rights groups called for “independent monitoring” of factories by impartial, well-trained observers, especially local NGOs that could gather information through off-site interviews with workers.
The fight over independent monitoring played out in large part through the construction of larger organizations to coordinate rule making and auditing. With activists arguing that companies should not be trusted to police themselves and with companies looking to lend credibility to their standards and protect their reputations, there were demands on both sides of the aisle for greater oversight. However, there were quite different views of what this should entail, and the highly contentious debates that ensued would ultimately lead to several different types of programs in the United States and Europe, each with different core constituencies. In other words, a number of different globalized localisms have sought to define “ethical” apparel.
Three main types of initiatives, defined largely by their core constituencies, have sought to define and oversee fair labor standards in the apparel and footwear industries. First, there are initiatives founded by coalitions of large brands and other stakeholders (e.g., NGOs and sometimes unions), such as the Fair Labor Association (FLA), Social Accountability International (SAI), and Ethical Trading Initiative (ETI). In various ways these initiatives sought to help brands improve conditions in their suppliers’ factories and to uphold the brands’ substantial investments in their corporate images. Because industry outsiders were involved to some degree in their creation, these initiatives adopted rules that pushed slightly beyond standard practices in the industry. This made many brands and especially retailers reluctant to join (Bartley 2009). Second, charging that these initiatives were too lax, NGOs and unions played more central roles in creating initiatives such as the Worker Rights Consortium (WRC) and Fair Wear Foundation (FWF). These set higher standards, with a particular emphasis on the rights of workers, not just their “protection.” Unlike others, these initiatives conduct their own monitoring and investigation, but they operate on a small scale within particular market niches. Third, industry associations sponsored their own programs—namely, the Business Social Compliance Initiative in Europe and the Worldwide Responsible Accredited Production program in the United States—that attracted a larger swath of corporate participants, including those with lower brand investments. They developed baseline standards with little input from industry outsiders. In addition, rather than being focused on protecting brand reputations, these initiatives have sought to give their members information about the compliance profiles of suppliers. As we describe below, these three models unfolded dynamically over time, with the first set of programs—that is, those negotiated with some “multi-stakeholder” input—emerging first.
President Clinton’s convening of the Apparel Industry Partnership in 1996 marked the beginning of a struggle to define the responsibilities of American brands and retailers. In the wake of the Kathie Lee Gifford scandal, in which child laborers in Honduras were found to be producing her line of children’s clothing, the Clinton administration called together a group of representatives from apparel and footwear brands, NGOs, and unions. Early talk of a “no sweat” label quickly fell by the wayside amid companies’ resistance to the “knotty issues” this would entail (Ramey 1996). For their part, labor rights advocates were reluctant to grant a simple endorsement of companies in a complex and largely nonunionized industry. The group did develop an overarching code of conduct, but disagreement quickly surfaced, especially over whether a “living wage” should be required and the meaning of “freedom of association” in places where unions were legally restricted (e.g., China).
The disagreements intensified as the Apparel Industry Partnership spawned the Fair Labor Association (FLA) to oversee compliance with this code. Several smaller companies had already left the partnership, leaving mainly large, brand-sensitive firms (e.g., Nike, Reebok, Liz Claiborne, and Phillips Van Heusen). UNITE and the Interfaith Center for Corporate Responsibility dropped out when the FLA was announced, charging that its monitoring amounted to “the fox guarding the henhouse” (Benjamin 1999). With a depleted set of NGOs, FLA leaders recruited universities, which were facing student protest. By 2001 FLA had begun its program of having accredited monitors visit a sample of member companies’ factories. It did not certify factories or produce a consumer label, but it did lend its approval to the compliance programs of members, befitting these companies’ interest in a form of “reputation protection” that would not overshadow their own brand images.
In the UK a somewhat similar organization, the Ethical Trading Initiative, was being formed, again with government support and appealing primarily to image-sensitive firms. But in this case labor unions and NGOs such as Oxfam and Women Working Worldwide were more influential, shaping many of the standards that would become the ETI “base code” (Schaller 2007). It garnered some participation by apparel companies such as Marks & Spencer and Levi Strauss, but the ETI also addressed a wider array of products, including tea, wine, and cosmetics.
Another brand-NGO coalition produced an initiative called Social Accountability International, to certify particular factories. The Council on Economic Priorities, a New York–based nonprofit focused on socially responsible investment and consumption, teamed up with fashion brand Eileen Fisher, German catalog retailer Otto Versand, Toys R Us, the auditing firm SGS, and several others to produce the SA8000 standard for factory certification. A representative of the International Textile, Garment, and Leather Workers Federation also supported the standard, but its US affiliate, UNITE, saw SA8000 as no better than the corporate-dominated FLA. The number of SA8000 facilities worldwide rose from eight in 1998 to more than one hundred in 2001 and to nearly three thousand in 2012. There is no product label, but factories can advertise their certification, and lead firms that support SAI, including the Gap, Timberland, and Tchibo, can advertise that involvement to consumers and, perhaps more importantly, to audiences in the socially responsible investment community.
These three organizations—FLA, ETI, and SAI—make frequent references to ILO conventions and ask companies to promote “best practices” for labor conditions in their supply chains. While the FLA largely ignored the conundrum of freedom of association in countries where independent unions were prohibited, ETI and SAI called for factory managers to support or allow “parallel means” of worker representation, such as worker committees. On the other hand, these programs did not require dramatic changes in the organization of production. Lead firms could produce more or less wherever they wanted and continue to demand low prices from their suppliers so long as they pushed for marginal improvements in working conditions. As we will see, this has greatly limited the impact of standards.
Labor rights activists challenged “checklist monitoring” and corporate-dominated initiatives by developing their own programs. Some groups that had exited the FLA joined with student anti-sweatshop activists and sympathetic scholars to create the Worker Rights Consortium (WRC), which focused on collegiate-licensed apparel. Enrolling universities as opposed to companies as its members, the WRC adopted a “fire alarm” model (McCubbins and Schwartz 1984), in which complaints from workers would trigger investigations by the WRC, drawing on local experts. Rather than “credentialing companies” and their claims of responsibility, the WRC would, as one developer put it, “credential workers” and their claims about injustice.5 Having successfully recruited a critical mass of colleges and universities, the WRC quickly became an important player in identifying violations of labor rights, especially those involving the repression of independent union organizing.
In Europe the Dutch affiliate of the Clean Clothes Campaign (CCC) was the key architect of the Fair Wear Foundation (Bair and Palpacuer 2012; O’Rourke 2003). Unlike the WRC, the FWF initially cooperated with industry, but the CCC’s insistence on substantive verification of factory conditions by well-qualified local groups led the representatives of large apparel retailers to leave the project (Fransen 2011). Some smaller apparel brands remained, though, making the FWF a somewhat unusual program: a “multi-stakeholder initiative,” in which the balance of power tilts toward nonindustry groups, including NGOs and unions.
On the other side of the aisle, apparel industry associations developed their own alternatives. The American Apparel and Footwear Association, whose members accounted for the vast majority of garments sold in the United States, created a shared code of conduct and a factory certification program: the Worldwide Responsible Accredited Production (WRAP) program.6 WRAP became formally independent of the trade association, but its personnel continued to have strong ties to the industry. At first glance WRAP’s standards appear similar to others, but they tend to offer greater flexibility. For instance, WRAP allows for exceptions to the standard regarding maximum days of work when “required to meet urgent business needs.” By setting a basic standard to which many factories can be certified, WRAP sought to provide assurances that would be useful for the less visible, but still large, lead firms in the apparel industry. As one developer put it, “We believe Nike has to be able to stand up and say ‘Nike stands for this and Nike is this and that.’ At VF, we don’t need to do that, because nobody knows who VF is. But what we need to do is to be able to tell Walmart, ‘You can buy from us and [be] confident about this stuff, because we’re buying your product in factories that have been certified.’”7
The Business Social Compliance Initiative (BSCI) is the main industry-driven initiative in Europe. It was founded by the Foreign Trade Association, a group of retailers, brands, and other importers in consumer products industries. The BSCI’s formation was fueled in part by the exit of Dutch companies from the Fair Wear Foundation (Fransen 2011) and by the collapse of a nascent partnership between companies, unions, and NGOs in Sweden (Egels-Zandén and Wahlqvist 2007). Although it was unions and NGOs that dropped out of coalitions in the United States, it was companies that did so in Europe. Like WRAP, BSCI’s goal is to coordinate auditing so that results from a single BSCI-approved audit of a given factory can be available to all member companies (Egels-Zandén and Wahlqvist 2007). It relies on auditors that have been accredited to grant SA8000 certification, but BSCI does not actually certify factories, and its expectations for factories are somewhat lower than the SA8000 standard. If a factory does not show improvement, BSCI “encourage[s] participants to reconsider their relations with that producer,” but it does not require them to do so.8 BSCI has become the main compliance initiative for European companies, including mega-retailers such as Metro (Germany), E. Leclerc (France), and El Corte Ingles (Spain), specialty apparel retailers (e.g., KappAhl, Esprit), and numerous other brands.
Codes of conduct and factory auditing/certification initiatives have certainly not transformed the apparel and footwear industries. Existing evidence suggests that they have had some meaningful but narrow effects on working conditions and the management of human resources. But the rights of workers have been less affected, and even on the issues where codes tend to be most meaningful, standards in many parts of the industry remain criminally low in an absolute sense.
Researchers generally have found that codes of conduct, and the auditing thereof, have led to some improvement in workplace health and safety (Barrientos and Smith 2006; Mamic 2004). For instance, looking at toy manufacturers in China, Niklas Egels-Zandén (2014) found that scrutiny from buyers led factory managers to provide health and safety education to workers. Examining the apparel industry in Bangladesh, Faisal Z. Ahmed and his colleagues (2014) suggest that despite a repressive political environment for workers, supply chain scrutiny has made factory managers slightly more sensitive to “worker welfare.” In the athletic footwear factories of Southeast Asia and China, the toxin Toluene had been widely used as a glue, but scrutiny from Nike, Reebok, Adidas, and New Balance have made it far less common. Officially it has been phased out, “except when it’s not,” as one compliance official for a major brand put it, admitting that it often works better than substitutes.9
Scrutiny from auditors and brands has also reduced, though not eliminated, the incidence of especially harsh, abusive treatment of workers. Early exposés showed managers screaming insults at workers or using physical punishment to induce greater productivity. In 1996 managers at a footwear factory in Vietnam producing for Nike ordered workers to run around the factory in sweltering heat as punishment, leading some of them to faint (Herbert 1997). Over time and with prodding from its buyers, the company that owned this factory, Pou Chen/Yue Yuen, has embraced a gentler and more formalized model of human resource management. Visiting a Yue Yuen factory more recently, we found an internal compliance staff with little tolerance for abusive supervisors, as well as a “counseling room,” hotline, and “worker welfare committee” for workers to lodge complaints.10 Such reforms are often cosmetic and rarely give real voice to workers. But they suggest that codes of conduct have shifted at least the first tier of suppliers to global brands away from intimidation as a management strategy.
On the other hand, improvements are certainly not guaranteed. Even brands that have audited suppliers closely have not necessarily spurred progress. Nike developed one of the most sophisticated auditing programs in the industry, but the vast majority of its suppliers’ factories—roughly 80 percent—failed to improve from 2001 to 2004, and some actually experienced a decline in their overall compliance rating (Locke, Qin, and Brause 2007). While codes of conduct have sometimes increased compliance with minimum wage standards and reduced arbitrary pay deductions (Barrientos and Smith 2006), one can still find suppliers to major brands paying below the legal minimum (e.g., Worker Rights Consortium 2013b). Furthermore, in many apparel-exporting countries the value of garment workers’ wages has decreased during the period that codes of conduct have been in place. In a recent study of fifteen major exporting countries, research found that in nine of these countries, the local purchasing power of prevailing wages declined from 2001 to 2013 (Worker Rights Consortium 2013a).11
When it comes to the rights and empowerment of workers, codes of conduct have had scattered and short-lived impacts at best. Industry-driven programs (e.g., WRAP and BSCI) have shown little interest in making respect for labor rights a necessary ingredient of socially responsible production. But even multi-stakeholder initiatives have largely failed as guarantors of labor rights. In their evaluation of ETI, Stephanie Barrientos and Sally Smith (2006) found that codes of conduct had made a difference for health and safety at work but essentially no difference when it came to workers’ empowerment. Retaliation against union supporters was common in many of the sites they studied, and “codes have done little to address this” (30). Codes “have had some superficial impacts on discrimination [in employment]. But they have had no impact on underlying patterns of employment based on gender, ethnicity, caste and religion” (30). In Cambodia the Better Work Program, sponsored by the International Labor Organization and International Finance Corporation (IFC) has worked to help factories comply with labor standards set by foreign buyers. It has resulted in a number of improvements in working conditions (Brown, Dehejia, and Robertson 2013), but outside the factory insurgent labor unions in the Cambodian garment industry have faced repression, as seen in the 2004 assassination of union leader Chea Vichea (Hughes 2007) and the 2014 police shooting of striking garment workers (BBC 2014).
In a few apparel and footwear factories, grassroots unions have effectively used the “brand boomerang” strategy to counteract resistance and repression from factory managers and local governments. In the best-known cases, independent unions at the Kukdong factory in Mexico and BJ&B factory in the Dominican Republic successfully gained collective bargaining rights after international campaigns led Nike to intervene in support (Anner and Evans 2004; Ross 2006). In another case, Liz Claiborne and the FLA played positive roles in preserving space for an independent union at the Choishin factory in Guatemala (Rodríguez-Garavito 2005; US/LEAP 2003). But the success of these campaigns depended on a rare confluence of factors in addition to codes of conduct—namely, grassroots organizing, strong cross-border activist networks, and leverage from trade agreements (Rodríguez-Garavito 2005; Seidman 2007; US/LEAP 2003). Furthermore, a number of factories in which unions gained recognition by “leveraging” codes of conduct later shut down, including BJ&B, Choishin, and factories where unions had gained ground in Indonesia, Thailand, and Sri Lanka (Egels-Zandén and Bartley 2014; Evans 2010).12
Recent industrial accidents (such as those mentioned at the beginning of this chapter) have provided a brutal reminder that even though occupational health and safety conditions have sometimes improved, serious threats have not been eliminated. In 2012 an apparel factory in Pakistan that had just been approved for SA8000 certification caught fire, killing nearly 300 workers who were trapped behind locked emergency exits and barred windows (Walsh and Greenhouse 2012). In Bangladesh a series of factory fires from 2010 to 2012 killed several hundred people in factories producing for—and audited by—Tommy Hilfiger, Walmart, C&A, and others. At one of these factories, Tazreen Fashions, auditors for BSCI and Walmart had noted problems, but the factory was still producing for a BSCI member (C&A) and as a subcontractor for Walmart (among others) when a fire killed 112 workers (Theuws et al. 2013). Sadly, the scale of these disasters paled in comparison to the 2013 collapse of the Rana Plaza building in Bangladesh, where more than 1,100 workers died. Factories within this complex were producing for British retailers Primark, Benetton, and others, and for Walmart just prior to the disaster (Greenhouse 2013). BSCI auditors, who had visited factories in Rana Plaza, were left in the awkward position of explaining that “audits do not include building construction or integrity.”13 In short, after nearly two decades of codes of conduct, factory auditing, and a variety of related projects, the global apparel industry bears strong resemblances to the period prior to the anti-sweatshop movement—and even to the period of the Triangle Shirtwaist fire.
The rise of codes of conduct would appear to make the apparel industry increasingly rule-governed. Rather than feigning ignorance or looking the other way, lead firms were beginning to present factory managers with long lists of what they could and could not do and sending auditors to police their behavior. Yet in many ways the implementation process and the industry have proven quite unruly. Specifically, three forms of unruliness help to explain why rule-making projects have had such a limited impact in this industry.
First, the auditors who were supposed to police factories in many cases turned out to be of poor quality or even corrupt. Early critics of “checklist monitoring” argued that factory auditing should be done by skilled observers who could gather trustworthy information from workers—by taking the time to conduct off-site interviews, for instance (Esbenshade 2004; Labor Rights in China 1999; O’Rourke 2002). Some specialized and dedicated nonprofit monitoring organizations did emerge (e.g., Verité in the United States, Impactt in the UK, Coverco in Guatemala). But as factory auditing became more common, the larger, lower-priced auditors (e.g., generalist firms with office around the world, such as Intertek, Bureau Veritas, SGS, and TÜV) came to dominate the market. In effect, factory auditing became commoditized—that is, treated as a low-cost, highly substitutable good—rather than becoming a profession. Specialized and nonprofit monitors still perform some audits and trainings, but the large generalist firms do most of the auditing for BSCI, SA8000, and WRAP, in addition to being hired by individual firms.
Despite oversight from accreditation bodies, the auditors from these firms have earned a reputation for being inexperienced, overstretched, and willing to turn a blind eye to problems when necessary to please factory managers. According to an auditor for one of the large firms in Vietnam, auditors are offered bribes by roughly 30 percent of factory managers, and others offer dinners and accommodations to try to curry favor.14 Auditors have also been caught tipping off factory managers about the timing of a supposedly unannounced audit.15 In China an undercover investigation found rampant corruption in Bureau Veritas’ factory auditing, including extortion by auditors and bribery by consultants who had been hired to guarantee that the factory would pass (China Labor Watch 2009).
The problems of factory auditing have become widely apparent over time. Exposés in Business Week and the Financial Times have shown how factory managers, especially in China, offer bribes, keep falsified records of wages and working hours, and coach workers to give the “correct” answer to auditors (Business Week 2006; Foster and Harney 2005). Practitioners have long recognized the cat-and-mouse game that even honest and dedicated auditors get trapped in: auditors chase the elusive “real” data, managers offer suspicious or partial records, and workers parrot answers that auditors suspect are coached. In response to these problems, some rule-making projects have lost faith in factory auditing. The ETI declared a “growing crisis in ethical trade auditing” (Ethical Trading Initiative 2006), and the FLA soon moved away from its original monitoring model, calling it an “inadequate tool to create sustainable change in working conditions,” in order to instead emphasize learning and “capacity-building” at the factory level (Fair Labor Association 2007).
Second, rule-making projects have done little to alter the sourcing practices of lead firms, which are a root cause of many factory-level problems. Even imperfect auditing can identify problems, but this does not mean they are necessarily resolved. Lead firms have demanded that suppliers comply with codes of conduct, but their sourcing priorities—low prices and fast delivery—send quite different messages. To the standard criteria of “quality, delivery, and price” that brands and retailers use to place their orders, “compliance,” in theory, has been added. But as one brand representative put it, “We both know that’s bullshit.”16 Some brands and retailers have well-staffed and well-meaning compliance departments, but these departments rarely have the power to shape the decisions of the production/sourcing departments. As assessors for ETI found, “Some suppliers believe one standard is operated by buyers who make commercial demands which require certain labor practices (such as long overtime). Another standard is operated by code of conduct managers, who deem the same practices as non-compliance” (Barrientos and Howell 2006, 9). In a study of one large brand at the forefront of CSR efforts, researchers found that more than half of the company’s current suppliers had not been approved by the company’s compliance department (with non-approval rates of over 85 percent in South Asia and East Asia). As these researchers put it:
While sourcing departments continue to squeeze factories on price, compress lead times, and demand high-quality standards, compliance officers visit the factories and document the problems but do little to change the root causes underlying poor working conditions. [One] auditor reported that “if [the sourcing department] has already sold the sample before I set foot in the factory, I know that we will give them business no matter what.” (Locke, Amengual, and Mangla 2009, 335)
In this situation the compliance efforts of brands and retailers are at worst an exercise in futility and at best akin to janitorial work. The role of compliance staff is to clean things up—that is, to spur minor improvements in factories that can produce the right product for the right price—rather than to identify factories with decent working conditions. For factories, improvements that raised costs or made it more difficult to produce large volumes in short time periods would certainly mean a severe risk of losing business.
Relatedly, factory managers often use subcontracting to meet deadlines and control costs, and brands’ scrutiny rarely reaches these second- and third-tier suppliers. Tazreen Fashions, the site of a deadly factory fire in Bangladesh, had actually been dropped as a first-tier supplier for Walmart, but it was continuing to work as a subcontractor, supposedly unbeknownst to Walmart officials. It is the complexity and poor scrutiny of subcontracting that has allowed the implementation of fair labor standards to coexist with especially harsh practices such as the resurgence of bonded labor schemes in the Indian garment industry (SOMO 2012). The Gap earned a reputation as a leader in the monitoring of its suppliers, but when journalists exposed bonded child labor at one of its Indian subcontractors, brand’s staff pleaded ignorance (McDougall 2007).
A third source of unruliness is the apparel industry’s geographical mobility. Lead firms have been quite willing—and able—to move their orders from one location to another to keep costs low. This has meant not just the decline of apparel manufacturing within affluent countries but also a series of rapid shifts from one set of developing countries to another. Places where labor conditions are improving typically have been abandoned for new locations. As described above, unions in Indonesia, the Dominican Republic, and Thailand have gained ground in some factories, only to see those factories, and often those countries, lose orders. In addition, researchers have shown that when brands and factory managers cooperate over long periods of time, they can improve working conditions (Frenkel and Scott 2002; Locke, Amengual, and Mangla 2009). But the mobility of the industry undermines these types of slow improvements. In essence, lead firms must restart their compliance efforts as they move to new locations.
It is not just mobility in the abstract that limits improvements; it is also the industry’s gravitation toward places with shoddy infrastructure and regulation that is poorly enforced or intended to constrain workers’ rights. Consider two countries where apparel exports have risen, Bangladesh and Pakistan. In Bangladesh the rapid buildup in the apparel industry came at the expense of building safety, because entrepreneurs built multistory factories, which are known to be hazardous. “It was designed as an apartment building, and now it’s a factory,” said one of the few engineering inspectors working there (qtd. in Srivastava and Shannon 2013, 2). As of the mid-2000s the Bangladeshi government employed just seventeen engineering inspectors to cover the entire economy (Miller 2013, 141). Furthermore, the political context in Bangladesh has privileged garment factory owners and marginalized their workers (Ahmed, Greenleaf, and Sacks 2014). In Pakistan government labor inspection has been dramatically scaled back (US Department of Labor 2012), and in some regions inspectors cannot check a factory within its first year of operation or without the cooperation of the owner (International Labor Organization 2011, 369).
It is in large part because of the industry’s migration to these locations that apparel production continues to be plagued by factory fires, industrial accidents, bonded labor, and other extreme conditions. Based on the industry’s history, some of these conditions could be improved over time, but that would likely be followed by a shift to new destinations. In this sense, severe forms of exploitation are built into the unruly logic of the global apparel industry, and rule-making projects have done little to alter that.
Admittedly, some companies and initiatives have gone further than others to temper the forms of unruliness discussed above. Alternative rule-making projects such as the WRC and Fair Wear Foundation have been all too aware of how sourcing practices and industry mobility matter, since these factors have undercut many of their projects. The FLA and ETI have tempered their expectations of auditing, shifting instead to a focus on capacity building at the factory level. SAI has launched several projects emphasizing dialogue and capacity building, although they have also allowed SA8000 certification to expand under questionable circumstances. A few brands, such as Marks & Spencer and Nike, have done more than others to link compliance and sourcing decisions. But these projects have a great deal to prove, since “capability-building” projects have sometimes been quite thin, as the next chapter will discuss.
Even with the growth of production in Bangladesh, China remains the world’s top apparel exporter by far, accounting for 37 percent of total world clothing exports in 2010.17 China also holds particular challenges for the implementation of codes of conduct: factories are staffed by migrant workers from rural areas who lack rights to government services in the cities where they work and are often marginalized in a country where the urban-rural divide is strong. Guarantees of freedom of association are difficult if not impossible to square with the government’s restrictions on labor organization. In looking more closely at the implementation of voluntary labor standards in China, we focus on two key issues: how rules regarding freedom of association have been interpreted and how attempts to certify the best factories have fared.
Most codes of conduct, monitoring, and certification initiatives ask suppliers to respect workers’ freedom of association, including their rights to form unions and engage in collective bargaining. But the Chinese government limits union activity to the official, state-endorsed All China Federation of Trade Unions (ACFTU). ACFTU officials are closely tied to the state, and factory-level representatives are usually selected by union officials and factory management. More than “management unions” elsewhere, ACFTU unions have served as a means for the state to control workers (E. Friedman 2013). Even while emphasizing regional variation in ACFTU autonomy, Mingwei Liu (2010) concludes that “because the Party-State supports the unions with the aim of regaining control over society rather than protecting worker rights per se, too much dependence on the Party-State may cause the unions to act completely in the Party-State’s interests and thus become its tool for the maintenance of social stability” (50). Outside of the ACFTU there are some NGOs focused on supporting migrant workers, but they are in a highly precarious position. Authorities can (and do) revoke NGOs’ registrations or harass their leaders if they come to close to “sensitive” issues, and even those NGOs that are able to find a space where they can do meaningful work must constantly and cautiously manage their relationships (Chang, Ngok, and Zhuang 2010).
How, then, can companies and standard-setting initiatives implement rules calling for freedom of association in China? Companies have most commonly addressed this contradiction by simply ignoring it—that is, by overlooking restrictions on unions and associations of workers in China. In a few telling cases, companies have sought other ways around the problem. The most ambitious effort was taken by Reebok in the early 2000s. In two footwear factories in South China, Reebok sponsored elections whereby employees could select their own ACFTU representatives. Although this experiment began with a great deal of promise, and representatives were elected, it ended with reprisals from factory management and ACFTU officials, infighting among representatives, and mixed signals from Reebok (A. Chan 2009; Yu 2008b). Subsequent union leaders were not elected, and conditions in one of these factories got significantly worse over time (P. Lee 2007). What began as a groundbreaking experiment ended as a reminder that collective representation could not be created out of thin air, especially in China.
While they do not go as far as elected union officials, ETI and SAI call for “parallel means” of worker representation where union rights are restricted. In addition, some brands have promoted “worker welfare” or “health and safety” committees as ways of giving Chinese workers some voice. However, evidence of sustained and empowered worker committees is scant. In one case that has often been celebrated, an active worker committee was formed in a factory producing for Timberland (Center for International Private Enterprise and Social Accountability International 2009; Huang 2008; Huang and Guo 2006). But attempts to locate this committee in 2010 revealed that the factory had lost its orders from Timberland, as well as the manager who had supported the committee. In SA8000-certified factories, committees may be extremely rudimentary, dominated by management, or completely absent. As one certification practitioner admitted, “Auditors don’t really understand the purpose of the committee,” going on to explain that they often simply ask managers if there is a worker representative.18 Adidas’ compliance staff has facilitated the election of worker committees at several of its key supplier factories in China, but a meeting with one such committee revealed that its role was quite limited—weighing in on the food in the factory canteen and activities in the dormitory, for instance.19
In contrast, outside of channels provided by brands, Chinese migrant workers have sometimes been quite demanding. Through a wave of strikes they have engaged in a version of what Eric Hobsbawm (1952) called “collective bargaining by riot.” Strikes surged in the late 1990s and early 2000s, initially among urban workers in state-owned enterprises. But it was not long before migrant workers in export-oriented industries were waging strikes as well (Lee 2007; Silver and Zhang 2009). Most take the form of quick “wildcat” strikes in response to delayed wages or other immediate grievances, but in a handful of cases striking workers have demanded rights, not just compensation (Butollo and ten Brink 2012; E. Friedman 2013). The kind of formalized worker representation promoted by apparel and footwear brands is a far cry from this kind of grassroots mobilization. When asked if a worker with a more activist orientation would be allowed to be elected to the committee, one brand’s compliance expert answered, “If it was a strike leader? Almost surely not.”20
While rule-making projects have proven ineffective in promoting freedom of association in China, they have been more successful in formalizing factory management systems. Often this takes the form of what Ngai-Ling Sum and Ngai Pun (2005) call the “institutionalization of paperwork,” whereby a supplier subject to a buyer’s code of conduct develops “elaborate managerial and audit/documentation systems to defend the [factory] against charges of infringing the code” (197). Most of the managerial changes that this entails are little more than window-dressing, but there do seem to be differences between Chinese factories that are subject to buyers’ scrutiny and those that are not. For instance, examining codes of conduct in two apparel factories in China, Ngai Pun (2005) found “no cases of ‘bonding’ workers by requiring deposits of money from them when they are hired. Nor did either company keep workers’ identity cards, which other enterprises in China often do to prevent them from quitting. Disciplinary penalties were replaced by a system of rewards and compensation” (108).
As the most credible factory certification effort, one might expect the SA8000 standard to be applied only to those factories with superior management systems, where even if workers voices are muted their treatment should be above average. But a closer look at SA8000 certification in China reveals that there is wide variability among SA8000-certified factories, in large part because of lax auditing. As discussed above, falsified records and careless, sometimes corrupted auditing had become common in China, including among auditors accredited to grant SA8000 certification. By some accounts, managers facing an all-or-nothing audit for certification are especially likely to hide the true information, since failing would lead to the loss of business.21 Practitioners within China suggest that both the international accreditors and SAI “have not really been thinking about the challenges here.”22 As a result, as one brand representative put it, “some factories are getting certified by just hiring a consultant to get them certified.”23 In addition, certification audits might be scheduled at a particularly low point in the production season (to meet working-hour standards) or with enough advance notice to temporarily improve workplace conditions.24 Factory certification in China boomed in the mid-2000s, but oversight was weak, leaving practitioners uncertain as to how many certified factories are truly above and below the standard.25
Interviews with several workers (conducted outside factories) suggest that some SA8000-certified apparel factories in South China are indeed above average. One apparel factory producing for Nike and a Chinese brand appears to have reduced working hours while also offering wages that are comparable to or slightly above those in other factories in the area. But this is not always the case. At another certified factory in the same area, this one producing for Billabong and other brands, working hours remained well above the legal limit and the SA8000 standard.26 Some SA8000-certified factories have turned out to have working conditions that were well below average. In one certified home furnishings factory a follow-up visit revealed child labor, as well as horrendous health and safety conditions, such as workers making candles by hand over dangerous, primitive gasoline-bottle stoves.27 The variation in SA8000-certified factories has meant that when certified and noncertified factories are compared, the average differences are often negligible (see Bartley and Zhang forthcoming).
In sum, as global rules have been put into practice in China, they have often become watered down, whether to accommodate the authoritarian political environment or because of weak oversight of auditing. As the failed experiments with worker representation suggest, there was no quick fix for labor rights when the local context did not support it. And as the growth of SA8000 illustrates, simple assurances about “good” factories and “responsible” production of apparel and footwear are, although not outright shams, not reliable indicators. Even though wages have increased over time in China’s major manufacturing areas, tellingly, these changes have little to do with codes of conduct, factory auditing, or certification. Instead, as a large body of research is finding, wage increases have been driven by strikes, the tightening of labor markets, and the strengthening of Chinese labor law (see Kuruvilla, Lee, and Gallagher 2011). Domestic factors, it turns out, have been more important than global standards in bringing about marginal improvements in the apparel and footwear factories of China.
The limitations of codes of conduct and factory auditing/certification may rightly leave conscientious consumers looking for alternatives. Furthermore, while the projects described above are mainly focused on labor conditions in the final manufacturing process, there are also reasons to be concerned about conditions further upstream in the supply chain—from water pollution and worker poisoning in the manufacturing and dyeing of textiles in China and India to forced labor in the cotton fields of Uzbekistan. Perhaps some emerging models of production in the apparel industry can provide more opportunities for conscientious consumers.
As labor costs have risen in China, some apparel manufacturers have partially “re-shored” production to the United States and UK (Bishop 2011). Fashion designers and upscale brands such as Eileen Fisher and Brooks Brothers have been able to allow consumers to bear the increased costs of manufacturing in domestic garment districts. But some other attempts to return to domestic production, particularly in mass-market goods, have foundered. The department store Dillards, for instance, quickly abandoned its re-shoring project, and the Pennsylvania company that was supplying it soon went out of business (Clifford 2013; A. Friedman 2012). While a “Made in the USA” label on apparel does suggest higher wages and a decreased chance of serious safety hazards (e.g., factory fires), it is hardly a guarantee of ideal working conditions. Immigrant workers in domestic garment districts in New York, Los Angeles, and elsewhere are often in a precarious position and have been subject to below-minimum wages, unpaid overtime, and retaliation against unionization efforts (Bonacich and Appelbaum 2000; Ross 2004). Some unionized apparel factories remain, but many of these shut down in the late 1990s.
The most prominent effort to promote domestic production has come from American Apparel, a vertically integrated retailer and manufacturer with a factory in Los Angeles. The company pays factory workers above the minimum wage and offers health insurance. It has at various points portrayed itself as “sweatshop free,” although it has recently focused its marketing on a “Made in Los Angeles” identity, as well as sexually provocative advertisements that many have charged with being porn-inspired and misogynistic. The company’s recently ousted CEO has also been charged with sexual harassment by a number of retail employees and has engaged in a number of explicit acts in the workplace. Despite its occasional attempts to claim a moral high ground, the company has also proven hostile to unions, fighting vigorously against early attempts to organize its manufacturing workers (Brooks 2007). Overall, American Apparel’s manufacturing process does differ from most of the industry in some respects, but its antiunion stance and regressive gender politics have made labor rights advocates extremely reluctant to endorse the company.
An alternative model that takes rights more seriously is the Alta Gracia factory in the Dominican Republic, where unionized workers earn a living wage making T-shirts primarily for the collegiate market. The story of Alta Gracia begins with the failure of the BJ&B factory in the same location, which produced baseball caps for Nike and Reebok. In the late 1990s anti-sweatshop activists in the United States exposed management abuse of BJ&B workers and repression of their efforts to form a union. With support from campaigns targeting Nike and investigation by the WRC, BJ&B workers successfully formed a union and engaged in collective bargaining in 2003 (Ross 2006). But this victory was short-lived. Within several years the Korean company that owned the factory and many of its buyers were shifting to factories in Vietnam, Bangladesh, and other Asian countries. The BJ&B workforce shrank to one-fifth of its previous size, and in 2007 the factory shut down (Clean Clothes Campaign 2007; Nova and Kline 2014).28
Out of the rubble of the BJ&B campaign, the WRC brokered an agreement with Knights Apparel, a US-based apparel company, to open the Alta Gracia factory at this location. Knights Apparel was a major manufacturer of collegiate-licensed apparel but did not have a strong brand image, and its CEO had developed a personal interest in social responsibility (Kline 2010). The factory opened in 2010, hiring some former BJ&B workers. It paid a living wage as calculated by the WRC and local experts—$115 per week, roughly three times higher than the legal minimum wage or the prevailing wage in the apparel industry in that region; and workers got cushioned ergonomic chairs rather than the commonly used metal stools (Kline 2010). Most notably, management had agreed to negotiate with a union, which workers quickly organized. The factory has navigated several production challenges and has succeeded in getting its products into university bookstores, as well as selling through the bulk wholesaler Ethixmerch. As of 2013 it had survived but was not yet turning a profit (Northam 2013).
Alta Gracia is only one small factory. The conditions that facilitated this project, including the prior unionization efforts and broader reforms in Dominican labor standards (Schrank 2013), make it difficult to replicate elsewhere. Nevertheless, it demonstrates that it is possible to produce apparel in developing countries and respect labor rights. Although labor rights activists have been reluctant to follow organic and fair trade advocates in seeking to build markets, the Alta Gracia project may be changing that (Nova and Kline 2014:274).
Finally, as seen in previous chapters, consumers of coffee, flowers, sugar, and chocolate have turned to fair trade certification as an alternative to the mainstream industry. Some cotton farms in developing countries are also part of the fair trade system (Bassett 2010), and apparel brands have begun to sell clothes made of fair trade–certified cotton. Note, however, that using fair trade cotton has no implications for the manufacturing process, despite what advertisements might imply. There is not yet a process in place to certify an entire fair trade garment, but Fair Trade USA has recently been working on one. Their system would certify both cotton producers and sewing operations so that a full garment could be labeled as Fair Trade. The challenges of adapting the fair trade model to wage earners in manufacturing operations are substantial, especially in a rapidly changing, highly mobile industry like apparel (Maquila Solidarity Network 2006). Nevertheless, Fair Trade USA pushed forward with a pilot project for garment manufacturing (as it has with other extensions of the fair trade model, as described in chapter 4). It appears that Fair Trade manufacturing will take certification to the SA8000 standard as its baseline while also adding a premium of up to 10 percent of the factory price of the item, which is to be distributed to a “Fair Trade committee” of workers (Fair Trade USA 2012). It does not appear that a union will be required for a factory to be certified. Overall, it is too soon to say how meaningful of an alternative Fair Trade apparel might provide, but our examination of worker committees and SA8000 certification should raise questions about whether they can provide a solid foundation for real alternatives. On the other hand, while the initiatives described above have focused narrowly on labor conditions in manufacturing, the fair trade model at least has the virtue of taking account of conditions further upstream in the supply chain, as in cotton farming.
The Guardian newspaper’s “Ethical Fashion Directory” is a telling indicator of the state of alternative production models in the apparel industry. Reflecting the proliferation of green consumer assurances, it shows items made with organic cotton, sustainable fabrics, and recycled and vegan materials. But when it comes to labor standards, the offerings are slim. It shows a handful of brands with at least some products that are labeled “Made in the UK,” and it shows a much larger number of brands that use fair trade cotton. But as one looks further into these companies’ manufacturing processes, guarantees about working conditions are either absent or quite similar to the assurances about codes and auditing that are given by mainstream companies. Even in the UK, the bastion of conscientious consumerism, the search for ethical fashion remains confusing.
On the heels of “green” consumer movements, anti-sweatshop activists sought to add “fair labor” to the list of conscientious consumers’ concerns. They “named and shamed” well-known brands and urged consumers in affluent countries to act in solidarity with apparel and footwear workers elsewhere. There is no doubt that anti-sweatshop activists succeeded in getting the attention of companies and consumers. But the rule-making projects that emerged have largely failed to alter the logic of the industry or offer meaningful alternatives to consumers.
In large part this is because of the constituencies that had the most power in those projects. Large, high-profile brands were able to steer monitoring and certification initiatives (FLA, SAI, and to a lesser extent ETI) to approaches that did not require dramatic changes in their operations, even though the involvement of nonindustry groups pushed these firms a bit further than their competitors. This push was absent in programs sponsored by industry associations (WRAP and BSCI), which have lower standards and weaker enforcement. All of these initiatives have relied on auditing practices that, while variable, have often been quite shoddy. Unions and NGOs did take the lead in some smaller projects (the WRC and FWF) that have become key voices pushing for stringent enforcement and serious remediation of problems. But their scope of influence is small.
Aside from these programs, some brands and retailers have developed significant internal compliance programs. But to a striking extent, the activities of compliance personnel have been decoupled from the companies’ sourcing programs, where low prices and quick delivery—and geographical mobility in search of these—have remained the priorities. In short, there has been a proliferation of rules for fair labor conditions, but the industry has remained unruly in many respects.
So what is a conscientious consumer to do? Consumers’ desires to identify “good” and “bad” companies runs into the problem that the leading apparel brands and retailers are more similar than different when it comes to labor conditions. Even those with sophisticated compliance programs operate within a highly competitive, mobile, and unequal structure of production. The desire to push product prices down and to spend on marketing, design, and distribution has remained powerful. Conscientious consumers might seek to reward firms that are likely to be slightly better than average (such as those that participate in the ETI, FLA, or SAI), but the magnitude of this difference must be kept in mind. One can look for alternative models of production, although as we have seen, domestic production and fair trade assurances are not necessarily reliable indicators of decent labor conditions in the apparel industry. One can seek to build collective consumer power in support of stringent standards, as university students have for the WRC. But concerned citizen-consumers must also look for other ways of promoting global labor standards, such as supporting trade policy that allows for the strengthening of labor rights and opposing trade policy that ignores or undermines labor regulation.
Some recent initiatives in the apparel industry suggest some tentative movement toward new approaches. First, the ILO-IFC Better Work program seeks to help factories in a given country to meet buyers’ codes of conduct and domestic legal requirements while still remaining competitive. Perhaps most importantly it seeks to bridge the gap between voluntary rule making and government labor inspection in developing countries. Originating from the project in Cambodia mentioned above, Better Work programs have been launched in Vietnam, Indonesia, Haiti, Jordan, Nicaragua, and Lesotho. Unlike the Cambodian program, these offshoots are not backed by trade agreements (which make Cambodian garments’ access to the United States dependent on improvement in labor conditions), and even in Cambodia repression of trade unions continued. Still, Better Work has the virtue of at least recognizing that governments must be part of any path toward decent work.
Second, in the wake of industrial accidents in Bangladesh, international labor rights advocates have convinced some lead firms to sign on to a binding Accord on Fire and Building Safety in Bangladesh. Signatory companies agree to support a new safety inspection and remediation program headed by international experts and to stop buying from factories that refuse to participate in the program.29 The number of signatory companies increased after the Rana Plaza collapse, reaching 140 by early 2014. These include European retailers and brands including Primark, Marks & Spencer, Carrefour, Kik, H&M, Adidas, and Benetton, as well as Philips Van Heusen, Abercrombie and Fitch, and eight other American companies. In contrast, many other American firms (including the Gap, Walmart, Target, and VF) have refused to accept a binding obligation, organizing a voluntary program on Bangladesh instead. The binding Bangladesh accord represents an intriguing model, although the baseline level of safety in Bangladesh remains well below many other apparel-producing countries. In another interesting move, the Walt Disney Company ultimately decided that Bangladesh (along with Pakistan, Belarus, Ecuador, and Venezuela) was too risky for its licensed apparel and toys to be produced there—a controversial but notable decision, since brands have rarely been willing to constrain their geographical choices.
These new approaches have some potential to improve on the failings of voluntary codes of conduct and checklist factory auditing. Still, unless the mobility and price sensitivity of the apparel industry is tamed to some degree, rules will give way to unruliness. Perhaps if conscientious consumption can translate into greater political engagement, especially with issues of trade policy and global labor regulation, the history of the apparel industry will not be doomed to repeat itself once more.