The Starbucks Solution: Can Voluntary Codes Raise Global Living Standards

Soon after protesters leafleted Starbucks stores because its Guatemalan coffee bean pickers earn less than that country's $2.50 daily minimum wage, Starbucks announced it would henceforth require growers to pay wages meeting "basic needs" of plantation workers. After NBC's Dateline filmed children working for Wal-Mart's Bangladesh garment suppliers, Wal-Mart adopted a contractor's "code of conduct." Following an expose that Levi Strauss contractors in Saipan had virtually enslaved imported Chinese women and paid them less than the legal minimum wage, the company vowed to do future business only with contractors whose workers are "fairly compensated . . . and not exploited in any way." A "toycott" of imported Chinese products provoked Toys "R" Us to adopt a code forswearing child labor. Reebok and Nike both now have codes for Third World workers, announced in response to Dutch and American protests (a recent Nation article noted that Nike paid Michael Jordan more for promotion in 1992 than it paid in total payroll for 30,000 Indonesian women). J.C. Penney, Kmart, Liz Claiborne, Nordstrom, Phillips-Van Heusen, Ford Motor, General Electric, AT&T, and many others also now have international sourcing codes, standards Third World companies must meet to qualify for orders from these American multinationals.

Following protests of its 1994 decision to extend China's Most Favored Nation trade benefits, the Clinton administration proposed that all companies manufacturing or sourcing abroad adopt voluntary "model business practice" codes with requirements like "avoidance of child labor," "respect for the right to organize," and a "safe and healthy workplace." (In response to corporate protests, a provision that American businesses pay a "fair wage" abroad was dropped.)

Considering the very modest content of NAFTA's labor and environmental side agreements, congressional votes to prohibit further negotiation of labor standards in trade pacts, and worldwide resistance to Clinton's proposal to inject labor standards into the World Trade Organization, voluntary codes may be the only way to go. Indeed, voluntarism is apparently the administration's second-best approach to a variety of frustrations. Stymied by declining enforcement resources, Secretary of Labor Robert Reich has jawboned retailers to police their own subcontractors for minimum-wage violations, and published a list of "trendsetters" who do so. In May, Clinton convened a televised seminar of corporate executives to heap praise on domestic firms that voluntarily provide benefits like child care or paternity leave. "I believe the power of example to change the behavior of Americans is enormous," Clinton said. "I want to . . . elevate the good practices that are going on . . . and hope that we can reinforce that kind of conduct." Earlier in his administration, the President was accused of focusing too exclusively on legislative bargaining and failing to take advantage of his "bully pulpit." No longer. Appeals for voluntary virtue are now the administration's answer to congressional reaction and gridlock. Reich says: "If government is going to do less, then the private sector will have to do more."

The issue of Third World labor standards takes on new urgency because of the globalization of commerce. Americans no longer have the luxury of expressing indignation about wages and working conditions overseas solely out of humanitarian concern, although dramatic inequality of incomes worldwide should be as much a part of public policy focus as domestic inequality. Foreign labor is also increasingly in direct competition with U.S. workers, so that when wages and working conditions overseas are excessively depressed, Americans find themselves importing the low labor standards along with the products. Equally important, Third World workers do not share in the fruits of their own labor. They are unable to afford consumption of either their own products or our exports, contributing to a looming crisis of international supply-side glut: lots of products with not enough people to buy them. If American participation in the new economy is to be based on promotion of exports, it is shortsighted and self-defeating to ignore the extreme forms of labor exploitation that inhibit the emergence of a worldwide consumer class.

Voluntary codes for multinationals' foreign operations are not an entirely implausible alternative to regulation: Recent codes have occasionally improved conditions of Third World workers. Levi Strauss, for example, reported it stopped doing business with 35 of its 700 or so worldwide contractors who failed to meet its new standards, and pulled out of China and Burma entirely. Some other firms now, in obedience to codes, use less child labor and may be less repressive towards union activists. In the absence of meaningful labor regulation or collective bargaining in developing nations, voluntary labor standards keep a spotlight on exploitation and may lead to demands for more effective action. The problem is that these codes are voluntary. Without adequate enforcement, codes can be mere public relations ploys, misleading consumers that workers' rights are actually respected in production.


Today's voluntary codes are partly inspired by the 1977 Sullivan Principles for investment in South Africa. Philadelphia civil rights leader Leon Sullivan had organized a 1960s boycott of retailers who discriminated in employment, and he subsequently established skills training centers for black youth in 125 American cities and 8 African nations. In 1971 when General Motors was pressured to diversify, Rev. Sullivan was invited to become the automaker's first black director. At his initial GM meeting, he challenged management by arguing in favor of a church-sponsored resolution demanding divestment of GM's South African operations.

In 1975, Sullivan met in South Africa with black leaders who asked him to stop pressuring American firms to leave the country and urged him instead to fight apartheid from within by making multinationals "agents of change." Despite opposition from other black leaders who supported divestment, Sullivan decided to give it a try. He developed six principles for American investment in South Africa: integration of eating facilities and restrooms; nondiscrimination in employment; equal pay for comparable work; training programs to move blacks into supervisory jobs; more black supervisors; and corporate support for improvements in blacks' housing, schools, and health facilities. Sullivan then demanded that GM and other firms sign on to these principles. Wearying of anti-apartheid demonstrations and boycotts, and fearful of possible legislation or executive orders to force divestment, 12 multinationals signed on initially and by 1985, 173 of the 300 American firms doing business in South Africa had committed. Subsequently, Sullivan amended the principles to include pledges to recognize black labor unions where they existed, to publicly oppose apartheid, and to pay a fair minimum wage. In the absence of a government standard, Sullivan defined this as the minimum needed to support a family of five, based on market-basket surveys of black townships by Johannesburg economists.

The anti-apartheid movement mostly condemned Sullivan's principles. Activists believed the principles lent legitimacy to apartheid by ameliorating its conditions while avoiding real pressure on the regime, which they believed could only come from divestment. And, they pointed out, the principles' potential for amelioration was itself limited. In 1984, Sullivan signatories' combined unskilled workforces were 99.6 percent nonwhite; managers were 5 percent nonwhite. "Equal pay" meant little if black and white South Africans rarely did comparable work. Desegregated restrooms in de facto segregated work areas were devoid even of symbolic effect. And while Sullivan signatories contributed nearly $300 million to public works (like improving African schools), this did relatively little to improve living conditions of 25 million black South Africans.

Subscribe to The American Prospect

By 1985, Sullivan concluded that maybe his critics were right. He announced that if apartheid wasn't dismantled in two years, he'd denounce his principles and demand divestment. It wasn't. He did. Though companies tried to maintain the principles without Sullivan, his conclusion that this alternative to divestment had failed hastened apartheid's downfall. Sanctions became inevitable.

In retrospect, the Sullivan Principles may have contributed more to the anti-apartheid struggle than activists acknowledged at the time. For while the codes did help multinationals deflect demonstrations, shareholder protests, and boycotts, it is improbable that major American firms were ready to succumb to anti-apartheid demands and divest in the late 1970s. It may be that political conditions for sanctions could not exist until multinationals had been given a chance to attempt gradual reform, and failed. Creation of protected zones (for black trade unions, for example) in some American-owned factories may also have chipped away at Afrikaner morale and will to resist. But whatever success we may now attribute to Sullivan's principles, the outlook for today's codes is less hopeful. Sullivan Principles were superior.

Sullivan had a single set of rules to which he demanded corporate assent. While the principles were sometimes compromised during Sullivan's corporate negotiations (he proposed, for example, that signatories commit to union shops, but dropped this when it became apparent that nonunion corporations in the United States would refuse to commit to unionization in South Africa), once Sullivan proclaimed a principle, companies could not separately dilute it. When, subsequent to the original proclamation, Sullivan several times toughened the principles—"raising the bar," he called it—the higher standards were uniform and signatory firms had to comply.

This made it possible to have a consistent monitoring system, understandable both to the companies themselves and to the anti-apartheid movement. Sullivan hired the consulting firm Arthur D. Little (ADL) to conduct the evaluations, with fees paid by a levy on each Sullivan signatory. ADL Vice President D. Reid Weedon rated each firm annually on a scale of I ("making good progress"), II ("making progress"), or III ("needs to become more active"), passing those who got "I" or "II," and flunking the others. Certainly, bargaining for favorable ratings was carried on by the most influential corporations, and firms like GM, Ford, and IBM passed with flying colors. Yet the ratings weren't all phony either. While Weedon's grades were based mostly on company questionnaires, covering quantifiable data like the number of black and white employees in each job category, he and his team also traveled frequently to South Africa to conduct inspections and to advise managements on compliance—how to desegregate facilities, implement a minimum-wage policy, bargain with a labor union, and so on. There were many failing grades (Carnation, Celanese, and W.R. Grace, for example), some South African managing directors were fired by U.S. parent firms when subsidiaries flunked, and some South African businesses raised wages to compete for labor with American companies that paid Sullivan's minimum wage. In 1980, company resentments about Weedon's demands (and about the burdens of responding to his annual 55-page questionnaire) became so intense that Sullivan signatories refused to pay fees. Weedon temporarily withdrew, but because public opposition to apartheid was sufficiently strong, the multinationals felt obligated to resume their payments, and Weedon resumed his role.


Today's post-apartheid codes don't have the advantage that identification with a single leader like Leon Sullivan could provide. Codes now vary, ranging from Sears's modest pledge to "comply with all applicable local" laws to Levi Strauss's commitment that its contractors' employees will be "fairly compensated and not exploited in any way." These promises have no external monitor, let alone enforcement. The administration has done little to follow up on its 1995 "model business practices"—by May of this year, there was still no official at the Department of Commerce (the cabinet agency charged with implementation) who was familiar with the program. But codes have become fashionable, and a plethora of nonprofit groups devote themselves to encouraging companies to develop and promulgate offshore labor standards:

  • The Council on Economic Priorities (CEP), publisher of a "shopping guide" for consumers on environmentally responsible companies, now lobbies corporate officials to develop more effective codes (particularly on child labor), and publishes an online evaluation of existing codes for institutional investors, soon to be included in its regularly published "Shopping for a Better World." But CEP does no independent evaluation of whether company codes are enforced. At its annual "Corporate Conscience Awards" dinner in June, CEP honored Starbucks for "its responsiveness to protests on behalf of Guatemalan workers," although Starbucks itself has not announced how it will calculate a wage level that assures the "basic needs" of peasant laborers.
  • The Interfaith Center on Corporate Responsibility (ICCR), which played a leading role in the anti-apartheid fight (coordinating money managers' demands for divestment at shareholder meetings), now disseminates "benchmarks" for corporations wishing to evaluate their codes, and for church-related pension funds wishing to base investment on corporate sourcing conduct. ICCR's model resolution includes a commitment that "our company should be in a position to assure shareholders that its employees are paid a sustainable community wage which enables them to meet basic needs, set aside money for future purchases, and earn enough discretionary income to participate in support of the development of small business in a local community." Needless to say, no company has yet adopted such a resolution. The Securities and Exchange Commission (SEC) has advised managements that they are free to deny votes on such resolutions because wage levels and personnel practices are not a "policy" matter, but simply "ordinary business" considerations. (Labor rights advocates hope recent Clinton SEC appointees will reverse this decision, but there's no evidence yet that they will do so.) Nonetheless, many corporations, feeling public pressure and wary of lawsuits, have permitted resolutions on less specific workplace standards to come to a vote.
  • Business for Social Responsibility (BSR) (with the incantation "running a company in a socially responsible manner isn't just the right thing to do, it makes good business sense") encourages member firms to adopt codes, and provides assistance in writing and promulgating them. BSR hopes to maintain an inventory of firms who adopt the administration's "model business practices," but no specific arrangements have yet been made.
  • The Investor Responsibility Research Center (IRRC) was founded in 1972 by the Ford, Carnegie, and Rockefeller Foundations to advise them on public policy issues where their shareholder votes could play a constructive role. IRRC is also a veteran of shareholder fights to force divestment in South Africa, and of a 12-year campaign of shareholder resolutions to force U.S. firms to adopt the MacBride Principles—commitments (also voluntary) to practice affirmative action policies for Catholic employees in Northern Ireland. The IRRC recently received a Ford Foundation grant to produce an inventory of contemporary corporate codes and evaluate whether they are an "effective mechanism for promoting worker rights and advancing workplace standards in developing countries." The grant, however, only includes funding for two research trips, one to El Salvador or Guatemala and another to Botswana, so the IRRC's evaluations will necessarily be imprecise.

Other groups have been playing roles as well. UNITE!, the garment workers union, has a provision in its contract with men's suit manufacturers that permits subcontracting only to plants that "provide a living wage defined as a specified market-basket of consumer goods." But public pressure to improve working conditions in developing nations is not now great enough to support demands that firms sign a common code, nor is there someone like Sullivan with the moral credibility to demand they sign on his terms. And while apartheid was a clear and easily identifiable evil, there is today no consensus among advocates about what a Third World sourcing code should contain. Agreement on avoiding child labor is relatively easy. But who will decide what is a "sustainable community wage"?


A few firms have their own inspectors who visit Third World factories, but most simply instruct buyers to inquire about labor standards when negotiating orders. And this usually entails nothing more than asking local management if it employs children and keeps fire exits clear, then dutifully recording the answers on compliance forms. Other buyers may walk through factories prepared in advance for their visits by herding children to another floor or into a restroom. A buyer in Bangladesh for Kmart and Wal-Mart, both of which have strong codes of conduct, boasts that he routinely certifies all factories as meeting company standards because "all the companies care about is the bottom line." A buyer for Liz Claiborne recently canceled Sri Lankan factory orders because poor lighting didn't meet the company's health and safety code; the buyer then contracted with a Hong Kong intermediary who in turn orders from the very Sri Lankan factory that the buyer himself just canceled. Now it's the intermediary's responsibility to verify compliance, not the Claiborne buyer's. A Wall Street Journal article last year ("Conduct Codes Garner Goodwill for Retailers, But Violations Go On") noted that even inspectors from Levi Strauss, one of the most progressive U.S. firms, "don't delve too deeply to avoid offending factory owners."

In three unusual cases, firms have agreed to outside monitoring. They illustrate the difficulties of a case-by-case approach:

Case One: Empresa Mandarin, an El Salvador garment plant constructed six years ago with U.S. Agency for International Development funds, pays wages of about 56 cents an hour and uses a private armed force to inhibit unionization.

Mandarin workers, many of whom live without running water or electricity in corrugated cardboard and tin shacks placed on vacant land along roadsides or polluted river banks, first attempted to unionize in 1993. In response, Mandarin fired all union leaders. In 1995 workers made another attempt. Notified by the Salvadoran government that a legal union had been formed, management locked out its 850 workers, some of whom were then attacked and beaten by security guards.

When the lockout ended, Mandarin workers were told to renounce unionization or be fired. About 150 who refused were then discharged. The remaining workers protested the illegal firings and the company again locked them out. A day later, management agreed to respect the country's labor code but then fired another 50 unionists.

Mandarin was producing at that time for The GAP, Eddie Bauer, J.C. Penney, Dayton Hudson (DHC), and other retailers. At DHC's shareholder meeting, an investment fund for several religious organizations demanded that the company require that its off-shore vendors respect employee rights to organize and to receive a "living wage." In response, DHC adopted a vague corporate policy to do business only with suppliers whose workers "have agreed to their compensation" and "who are not exploited." The GAP also adopted a code requiring vendors to "utilize fair employment practices" and prohibiting doing business with those who use "physical coercion as a form of discipline." J.C. Penney's rules demand "strict compliance with all applicable laws and regulations of the countries of manufacture."

Despite these codes, conditions at Mandarin didn't change, although inspectors from The GAP, Eddie Bauer, and Liz Claiborne inspected the Mandarin plant three times in 1995 and claimed they could find no violations. So the National Labor Committee, a project of UNITE!, took two fired Mandarin workers on a public speaking tour across the United States. The U.S. retailers subsequently announced they would no longer do business with Mandarin, but the Labor Committee, focusing on The GAP, intensified its campaign. It demanded that The GAP continue to outsource to Mandarin and other Salvadoran suppliers and that it use its relationship with its suppliers to pressure them into honoring The GAP's conduct code (including rehiring Mandarin unionists) and agreeing to independent monitoring to verify compliance. Last December, succumbing to demonstrators' demands during the Christmas shopping season, The GAP agreed, though it restricted the role of independent monitoring to El Salvador and neighboring Latin American countries. The Labor Committee, associated church groups, and GAP management chose the "Human Rights Ombudsman Office" in El Salvador as the monitor. To date, however, Mandarin claims it has not received enough new orders from The GAP to justify rehiring the workers.

Case Two: After several years of a public campaign by U.S. activists protesting child labor, Senator Tom Harkin and Representative George Brown proposed legislation in 1994 to prohibit imports manufactured by children. Meanwhile, the AFL-CIO assisted the organization of an independent union of garment workers in Dhaka, Bangladesh. The prospect of Harkin's bill, together with the union's threat to publicize child labor to American consumers, so frightened the Bangladesh Garment Manufacturers and Exporters Association that it negotiated a Memorandum of Understanding with UNICEF and the International Labor Organization (ILO) to end employment of children under age 14. Signed a year ago, the agreement requires each manufacturer to subsidize schools for previously employed child workers, to pay stipends for those who leave factories to attend such schools, and to permit the ILO to "devise a labour inspection system in co-operation with the Government of Bangladesh." To date, however, no stipends have been paid. While the ILO has made some factory inspections (and found children working), no mechanism has been established to correct or report the violations. And child labor is but a minor aspect of corporate code violations entailed in Bangladesh subcontracting. Many adult laborers work 17-hour shifts six and seven days a week, earn considerably less than Bangladesh's legal minimum wage (about $25 a month), and are denied compensation for overtime. ILO monitors are not chartered to inquire about wages or working hours, only about children's ages. Since the signing of the Memorandum on July 4, 1995, Bangladesh has often been paralyzed by political strikes. Continued crisis may destroy the nation's garment industry, preventing any real test of the efficacy of the novel agreement, even for its limited purpose of regulating child labor.

Case Three: German human rights groups have protested the import of carpets made by children in Indian, Pakistani, and Nepalese factories. Many of these children are bonded laborers, sold by parents or kidnapped by labor contractors. Joining with human rights groups in India, the German activists formed the Rugmark Foundation, which issues a label to certify, based on a foundation inspection, that carpets were produced without child labor. Inspections are now ongoing, although the huge number of child workers make large-scale evasion probable. A far greater problem for the Rugmark Foundation is how to advertise the label to German, Swiss, and American consumers; unless buyers discriminate in favor of carpets with labels, Rugmark will lose whatever credibility with manufacturers it now possesses.

These three lonely examples of code enforcement illustrate how difficult the current piecemeal approach will be. Each depends on public attention and consumer pressure focused on a single country or distinct part of the broader labor standards constellation. But there is a limit to consumers' ability to juggle multiple boycotts, while code violations by distinct companies or in distinct countries are too numerous to count. Separate public campaigns around several simultaneous violations will inevitably conflict and confuse. Rev. Sullivan had a difficult enough time maintaining pressure on investors in a single country to abide by a single code. Today, where each firm's code is unique, where monitoring rights are won only to remedy discrete and highly publicized violations, it is difficult to imagine how threats of future consumer or shareholder power can be maintained. Without greater coherence, corporate codes are unlikely to significantly influence Third World labor standards.

And what if, even in these circumstances, some battles are won? Imagine, for example, that the ILO-inspired Bangladesh agreement makes garments sourced in that nation truly child-free. The AFL-CIO, National Consumers League, and others who originally promoted a boycott of Bangladesh garments will have to decide whether they should now promote the purchase of these products—no longer made by children but by adult workers paid sub-minimum wages and denied the right to organize. Unless labor rights advocates promote purchase of products made in compliance with the child labor agreement, threats to compel concessions in future disputes will be hollow. Even where labor standards are honored by Third World contractors, wages in developing nations will for many decades continue to undercut those of American workers. American unions and their allies who are now in the forefront of the movement for international labor standards have not yet confronted the implications of possible victories. Implicit in any boycott is a commitment to purchase once demands are met. Are activists prepared to promote purchases of goods that many unionists will continue to regard as "unfairly" manufactured?


Those now encouraging voluntary corporate standards acknowledge the inevitability of code corruption without independent monitors. But without militant public pressure for effective Third World labor standards, American corporations have adamantly refused to consider external enforcement, while groups like CEP, ICCR, BSR, and IRRC, committed to nonadversarial corporate encouragement, can do little but gently wonder aloud how codes will be enforced. Reid Weedon, now retired from Arthur D. Little, was recently approached by several public pension funds exploring his willingness to establish a Sullivan-like monitoring system for firms in which the funds invest. Weedon is intrigued by the prospect, but there is no evidence that pension funds yet have the determination or the votes to force this mechanism on managements.

Meanwhile, Heather White, a onetime outsourcing agent for apparel companies, has founded a consulting firm to verify compliance with corporate codes. Her business plan for VERITE (Verification in Trade and Export) includes hiring human rights activists in Third World countries to inspect plants and report on code violations. The contract she now pitches to multinational corporations includes a provision that permits VERITE to publicize uncorrected violations six months after they were brought to corporate attention. White has the right idea, and she's highly regarded by groups like Business for Social Responsibility, but none of BSR's member firms have expressed serious interest in hiring her for this purpose.

But opportunities may develop. Last month, after embarrassing publicity about her private label garments manufactured in domestic sweatshops, as well as by children in Honduran plants, Kathie Lee Gifford announced she would hire an "independent" monitor for factories producing her apparel line for Wal-Mart. Wal-Mart itself then announced for the first time that it now "is considering" hire of "independent" monitors. Gifford held a press conference with New York Governor George Pataki as the latter pledged to introduce legislation banning the sale of garments manufactured by underpaid workers. New York, however, is in no position to dispatch inspectors to thousands of Third World sweatshops now feeding New Yorkers' hunger for fashion. If public pressure is maintained, and legislation is actually adopted, some monitoring operation will have to be created.

As production globalizes, ineffective labor regulation makes nongovernmental standards the only truly viable alternative. Multinationals rush to publicize their codes to deflect bad publicity, while church and labor activists, along with foundation and pension fund investors, nearly trip over each other to be helpful. If the "era of big government is over," these private initiatives to develop a worldwide civil society must proceed. But without fear of public and consumer pressure, corporations will have little incentive to make codes real. The success of these efforts to improve labor standards depends on mobilized outrage when they fail. Today, the exercise is too friendly, and lacks the uniformity, the rigor, and the external monitoring needed for success. But it's not a bad start.


illustration by Dan Yaccarino


You may also like