In the spring of 1992, a group of unionists gathered in a back office of the National Federation of Salvadoran Workers (FENASTRAS) to discuss their latest union campaign. The room was filled with excitement. Several months earlier the government and guerrilla combatants had signed a comprehensive peace accord that put an end to twelve years of violent civil war. After years of labor repression, workers appeared to be overcoming their fears. As high-ranking guerrilla commanders safely entered the capital city from their mountain hideouts or from exile in Managua or Mexico City, it seemed to workers that they too would be able to more freely exercise their rights.

FENASTRAS hoped to exploit the new atmosphere in the country by attempting to unionize workers in the growing export manufacturing sectors. At one factory, labor activists organized a protest event at the entrance of a facility. The factory owners responded in their customary way: They called the police and told them to come and immediately deal with the labor protest. The police arrived promptly, yet instead of dispersing the protesting workers with rifle butts, tear gas, and arrests—as had been customary during the war—they began directing traffic around the protesting workers.

Back in the union office, as news of the police action spread, the sense of excitement grew; the police did not take sides in the conflict, but rather came to direct traffic! A new era for organized labor had arrived in El Salvador, or so it seemed. Soon after the protest event and the same day that activists provided the Ministry of Labor with the list of workers that was needed in order to legalize their union, management began dismissing workers on the list. Indeed, the dismissals happened so quickly that a Ministry of Labor representative would have had to call or, more likely, fax the list of names of workers who had joined the union to the factory owner.

Something had indeed changed in postwar El Salvador. The new government had liberalized trade, welcomed foreign investors, and encouraged the restructuring of industries in order to compete in the global economy. And, while curtailing more violent forms of repression, it worked with employers to weaken unions in the hope that this would make the country more attractive to companies looking for low labor costs and little labor unrest. To survive, unionists would have to develop new forms of solidarity.

Across the Americas in Brazil, unionists were facing their own problems related to economic globalization. In an earlier era, the powerful metalworkers' union in the industrial suburbs of São Paulo known as the ABC Region had displayed a capacity to shut down an important segment of the national economy and contribute to a return to democratic rule. The union became one of the strongest industrial unions in Latin America. It had well-trained cadre, top leadership posts in the national labor confederation, and strong political influence through its ties to the Workers Party (PT). The union, which was once led by Luiz Inácio “Lula” da Silva, also commanded a militant and committed membership base.

So when Ford announced in 1999 that it was dismissing 2,800 workers at its São Bernardo plant, the workers assumed that they could do what they always did: show their unity and strength by organizing a powerful strike until the automobile giant was forced to revoke its decision. But the leader of the union realized that the time-honored labor strategy of striking might be losing its punch in the globalized economy. Under President Fernando Henrique Cardoso (1994–2002), the government had dramatically reduced import tariffs and domestic content requirements in the auto sector. Now a disruption in production in one country could quickly be picked up by another country. More troublesome still, the auto industry was facing a crisis of overproduction. Looking out at the huge stockpile of unsold cars in the Ford lot, the union president, Luiz Marinho, observed uma greve é pouco para responder (a strike is a small response). Like their Salvadoran colleagues, they too would have to develop new strategies.

Industrial Restructuring, Domestic Institutions, and Labor Identities

The challenges facing Salvadoran and Brazilian unionists are rooted, in part, in a dramatic restructuring of the global economy. The recent transformation of industry in Latin America has its roots in the 1970s oil crises and the breakdown of the Bretton-Woods fixed-exchange rate system, which led to the collapse of centralized, Fordist mass production systems.Footnote ¹ In an attempt to regain profitability, corporations developed complex systems of subcontracting and global networks of outsourcing for primary material, component parts, assembly production, services and distribution.Footnote ² In Latin America, this transformation was manifested in a transition from Import Substitution Industrialization (ISI) to export-oriented manufacturing. Rather than promoting domestic production for the domestic market by limiting foreign exports through high tariffs and domestic content requirements, governments began encouraging goods produced via global networks for exports, or, as in the case of the Brazilian auto sector, for the domestic market but now able to compete with an array of imports.

Market-oriented reforms engendered a deeper transformation in the structures of industry, or what Gary Gereffi and his collaborators refer to as “Global Commodity Chains (GCCs),” which consist “of sets of interorganizational networks clustered around one commodity or product.”Footnote ³ The commodity chain approach emphasizes the links between segments of the production process, distribution and consumption across the global economy. Gereffi makes a distinction between what he terms “producer-driven commodity chains” and “buyer-driven commodity chains.” Buyer-driven chains, which include apparel and footwear industries, “refer to those industries in which large retailers, brand-named merchandisers, and trading companies play the pivotal role in setting up decentralized production networks.”Footnote ⁴ For Gereffi, producer-driven chains, which include the auto industry, refer “to those industries in which transnational corporations (TNCs) or other large integrated industrial enterprises play the central role in controlling the production system.”Footnote ⁵

In Mexico, the Caribbean and Central America, buyer-driven GCCs dramatically expanded through Export Processing Zones (EPZs). In these large industrial parks, factories assemble products under contract for multinational firms and export them to American and European markets. There are an estimated two million workers in EPZs in Mexico, the Dominican Republic, and Central America. In most of these countries, EPZ production has surpassed traditional manufacturing in both the value of production and the number of workers employed. Employers prefer young women to perform the intense and monotonous work that often entails a work week of seventy hours or more. Burnout is common and has resulted in a very high turnover rate, factors that contribute to weaker worker identities and less stable labor unions. El Salvador's apparel industry is a case in point.

In South America, producer-driven GCCs, like the local automobile industry, also have been transformed. Gone are the huge production facilities that made cars from nuts and bolts to finished vehicles. Now not only are the small parts outsourced, but larger components are also produced by independent firms, only to be inserted into the vehicle on the final assembly line. Automobile production facilities are increasingly limiting their operations to pressing, assembling, and painting vehicles. Automation has further reduced the size of the workforce, as robots insert components and tighten the bolts that hold them in place. In 1980, Volkswagen in Brazil employed 37,467 workers in its plant outside São Paulo; in 2002, Ford began operations at its modern plant in Bahia with 3,300 workers.

These processes of industrial restructuring have taken place in a range of political and institutional contexts that have also affected workers and influenced labor activism. The historically high levels of unemployment and underemployment in Latin America have ensured that the market mechanism was unlikely to drive up wages. As a result, labor commonly pursued its demands by targeting the state.Footnote ⁶ Labor learned that it was more effective to take political action by demanding that the state increase the minimum wage or improve benefits (i.e. pensions, healthcare, vacation and severance pay regulation, etc.) than to attempt to employ market pressure by withholding labor at the point of production (i.e., striking). Thus, state institutions were always of critical importance for labor.

In the 1980s and 1990s, democratization and, in some cases, international pressure, led to reforms in collective labor laws in Latin America that were considered union-friendly.Footnote ⁷ Yet many of these reforms were relatively minor and did not provide labor with the needed protection for the new context created by the dramatic expansion of international production regimes.Footnote ⁸ In Central America, while it became easier to form unions, the threshold for collective bargaining remained high. This resulted in a proliferation of small unions with limited ability to improve wages and working conditions. In Brazil, constitutional reform was favorable to labor in that it allowed for the unionization of public-sector workers, but the reforms did not alter other aspects of the old corporatist system. The law continued to favor municipal level unionization and collective bargaining. For autoworkers, this did not matter as much in the 1970s and 1980s, since the industry was concentrated in one region of the country, the industrial suburbs outside of São Paulo. Yet in the 1990s, as the auto industry became dispersed throughout the country, the legal obstacles to national unionization and bargaining proved particularly deleterious for labor. For example, when Ford established a new factory in Bahia, instead of the workers coming under the single national contract already negotiated by the union at Ford in São Bernardo, workers would have to establish a new union in Bahia and negotiate without the advantage of the leverage provided by the older and stronger union in São Bernardo.

The combination of industrial restructuring and inadequate reform of labor-relations institutions created an enormous challenge for labor unions in Latin America's manufacturing sector. Old union strategies were no longer effective, and unionists were pushed to experiment with new forms of activism. The difference in the two forms of global commodity chains outlined by Gereffi helps to explain variations in new forms of labor activism. Since control lies with retail and brand-name companies in buyer-driven chains, labor unionists will be more likely to form alliances with nonlabor, consumer-oriented organizations through international activist networks similar to those described by Margaret Keck and Kathryn Sikkink.Footnote ⁹ Keck and Sikkink's focus on principled-values applies when studying workers in buyer-driven chains since these chains involve low-end sectors such as apparel, where workers often face sexual harassment, physical hardship, and below-subsistence wages. That is, their demands can be framed in terms of basic human rights that resonate with broader publics (not just labor activists). Yet internationalism, particularly among conservative unionists, is not always an automatic response that follows from changing structures of industry. Decades of political distrust fomented during the Cold War led some unionists to shun alliances with foreign activists and pursue nationalist pacts with employers.

In producer-driven chains, on the other hand, control rests with the transnational producers. Because of this, some unionists were inclined to focus their pressure on the production end of the commodity chain through international labor alliances. For example, Brazilian auto unionists need to put pressure on the Ford headquarters in the United States and the Volkswagen headquarters in Germany. To do this, they need to work with unionists in those countries more than with consumer activists. And since producer-driven chains involve high-end sectors where wages tend to be higher and human rights abuses less common than in the apparel sector, alliances with women's groups and human rights organizations will be less common. Yet, as in the apparel sector, where conservative unionists are not inclined to trust international activist alliances, conservative auto unionists may turn to pacts with employers to resolve their demands. Indeed, the greater stability of the employment relationship in the auto sector and the higher productivity (value-added generated per employee) allow for more sustained forms of cross-class collaboration.

What this suggests is that new labor strategies do not follow mechanically from new economic and political structures, but rather are filtered through historical experiences and ideational influences. As Robert Lieberman notes, what matters are “the meanings, interpretations, and judgments [actors] attach to events and conditions; and their beliefs about cause-and-effect relationships.”Footnote ¹⁰ Historical experiences are equally important. Charles Sabel explains, “Early choices naturally constrain later ones. Alliances create patterns of trust, fuse interests, reinforce or undermine organizations, and provoke counter-reactions by those they exclude.”Footnote ¹¹ During the course of the late nineteenth and twentieth centuries, anarcho-syndicalism, communism, populism, anticommunism, social democracy, and neoliberalism have influenced how unionists have defined and pursued their goals.Footnote ¹² Unionists adhering to radical ideologies were often the victims of state repression, which reinforced their distrust of the state. Unionists of a more conservative orientation, who had historically benefited from corporatist ties to the state and employers, tended to maintain a trust in cross-class collaboration despite larger economic and political transformations that greatly eroded the corporatist compromise.

The end of the Cold War lessened, but did not eliminate, the ideological differences among labor unionists. This was particularly true in Latin America, where the end of civil war and dictatorship meant that labor organizations no longer had to situate themselves between extremes. Yet differences in labor worldviews remain palatable. Left-oriented unionists tend to reject or at least distrust the capitalist status quo and seek significant social transformation. As the case studies in this article will illustrate, their class-based identity makes them more open to international alliances with labor and other social activists. In contrast, conservative unionists tend not to question the capitalist status quo but rather seek to find the best way to work within it. They adhere to a class-collaborationist world-view, and (at best) seek modest social reforms. These unionists tend to be nationalistic and distrustful of foreign unionists. When traditional union strategies do not work, they are more likely to seek alternative domestic strategies, including pacts with employers.

These relationships are illustrated by examining how labor unionists in the apparel export sector in El Salvador and the automobile sector in Brazil constructed new forms of labor activism in response to international industrial restructuring.

El Salvador

The challenges faced by the Salvadoran footwear and apparel workers depicted at the beginning of this article have their roots in a new economic model developed at the end of the country's civil war. The Salvadoran government, with an initial 26 million dollars in funding from the US Agency for International Development (USAID), promoted export-oriented manufacturing by building industrial parks that provided renewable exemptions on export taxes, import tariffs, and municipal taxes. Not surprisingly, the sector mushroomed. Jobs in the sector jumped from 3,000 in 1991 to 86,000 in 2000, and the sector's contribution to industrial GDP went from 4.3 percent to 11.2 percent.Footnote ¹³ By 2002, the sector generated 475 million dollars in value added,Footnote ¹⁴ surpassing export earnings of coffee, which had been El Salvador's most lucrative export for over a century.Footnote ¹⁵

Salvadoran state entities, which were promoting the system, acknowledged that it was based on extreme levels of competition that resulted in “cannibalism” among countries in which each country “tried to benefit as a result of geographic location, fiscal incentives, trade tariffs, and other aspects that make a country more attractive.”Footnote ¹⁶ The resulting low-paying jobs had specific demographic characteristics. In El Salvador, approximately 83 percent of the workforce was female, and 83.4 percent were thirty years old or younger.Footnote ¹⁷

The owners of these industrial parks advertised the low wages—paid to workers and mid-level managers alike—in order to attract foreign investors to the country. One EPZ website announced: “The minimum wage in El Salvador is an unbelievable USD 0.69 per hour. An experienced, college-educated, English-speaking middle-level manager earns from US$12,000 to 15,000 annually.”Footnote ¹⁸ The structure of the sector forced investors to keep wages low. Since multinational apparel companies provided local manufacturers all the major production inputs (fabric, thread, buttons, zippers, etc.), labor costs constituted the largest of the production costs. Successful competition thus entailed the maintenance of low compensation.

Keeping wages low also entailed curtailing the ability of workers to exercise their right to freedom of association. The use of intimidation, blacklisting, and dismissals to prevent unionization is well-documented.Footnote ¹⁹ These methods have proven to be effective. In 2001, government data indicated that while the manufacturing sector had a unionization rate of 9.83 percent, the unionization rate in the apparel export sector was 0.59 percent.Footnote ²⁰ My own survey of the apparel export sector in El Salvador found that unionization in the sector never surpassed 3.79 percent of the work force.Footnote ²¹

Yet, EPZ structures tell only part of the story. In the 1990s, the union movement was already weak and divided when industrial restructuring took place due to a long history of state repression of labor. In the late-nineteenth and early-twentieth centuries, the geographic concentration of coffee production and the limited incursion of foreign investors facilitated the formation of a strong, coherent national elite.Footnote ²² At the same time, limited industrial development and the seasonal nature of the coffee harvest engendered a weak labor movement.Footnote ²³ Unlike mining or banana production, most workers on coffee plantations are only employed during the short harvest season, which has limited their ability to develop a strong workers' consciousness and organizational ties with their coworkers. When labor did attempt to organize, such as in 1924 when unionists formed the Regional Federation of Salvadoran Workers (FRTS), the state acted to ruthlessly destroy these organizations.Footnote ²⁴

Intolerance of worker organizations was particularly extreme in the countryside. A peasant uprising in 1932 was brutally repressed. In its aftermath, the military killed an estimated 10,000 peasants, mostly in the western region of the country.Footnote ²⁵ Fifty years later, when the country descended into civil war, the collective memory of the 1932 massacre was still so strong that opposition activity in the western region of El Salvador remained limited.

During the rule of General Hernández Martínez (1932–1944), as James Goldston and Holly Burkhalter explain, “labor unions were outlawed on the grounds that they were inherently subversive and destabilizing.”Footnote ²⁶ The dictatorship was followed by a period of reforms. Newly-enacted labor laws combined extensive individual labor rights protections with limited collective rights.Footnote ²⁷ The system gave the state significant power to grant or deny legal recognition to unions, to authorize strikes, and to control collective bargaining. While most urban workers in theory had the right to organize, the law detailed how unions should be formed, what leadership positions should be filled, and how members must elect their leaders.

It was not until after the civil war (1980–1992) that the labor law was significantly modified. In terms of collective rights, the number of workers needed to form a union was reduced to thirty-five for any type of union.Footnote ²⁸ Yet, while it became technically easier for workers to form unions, it remained just as difficult to oblige an employer to negotiate a collective contract.Footnote ²⁹ As before, fifty-one percent of a work force needed to join the union. In the three years following the reforms, thirty-seven new unions and three new federations were formed, yet the national unionization rate remained unchanged at approximately ten percent of the workforce in the formal sector. Except for the construction sector, which was experiencing a postwar boom, the number of collective contracts and workers covered by them had declined significantly since the enactment of the reforms. In 1992, 102 contracts covered 39,667 workers. By 1995 there were only 79 contracts covering 37,201 workers.Footnote ³⁰

More troublesome still was the dramatic underfunding and understaffing of the Ministry of Labor. In 1995, there were only thirty workplace inspectors for the entire country.Footnote ³¹ There were also only fourteen health and safety inspectors with similar low salaries in the country.Footnote ³² To compensate for the low salaries, staff at the Ministry of Labor often compiled and sold lists of unionists' names to employers to facilitate blacklisting.Footnote ³³

The left-oriented labor movement's response to these challenges was filtered through a long and often violent labor history in El Salvador. In the 1920s and early 1930s, communist unionists adhered to the class conflict thesis of the Communist International and participated in the failed 1932 peasant insurrection. Afterwards, and again following the lead of the Communist International, Salvadoran Communists adopted the “united front” and, subsequently, “national antifascist front” strategies based on the belief that the more imminent threat to world revolution was fascism, not capitalism.Footnote ³⁴ They maintained this collaborative approach and a belief in electoral politics into the 1970s. Yet the Cuban revolution inspired the radicalization of a large segment of the union movement, which was simultaneously influenced by liberation theology.Footnote ³⁵

These unionists rejected the class-collaborationist position and the belief that a “bourgeois revolution” had to precede a “worker-peasant revolution.” They also pursued radical tactics such as sit-down strikes and mass street demonstrations that sought to bring down governments. In El Salvador, this union tendency gained prominence when sixteen unions walked out of the progovernment General Confederation of Trade Unions (CGS) to form the National Trade Union Federation of Salvadoran Workers (FENASTRAS). The federation, which in the 1980s became the largest labor federation in the country, prioritized political demands such as an end to political repression over more immediate economic demands such as wage increases.

Some members of the leadership of this growing “new” left union movement developed ties to the emerging guerrilla organizations.Footnote ³⁶ The combination of radical ideas, the successful example of the Cuban Revolution, and a perception of the country's history of dictatorship and labor repression, led this group of unionists to believe that the system needed to be dramatically transformed by combining increasingly radical labor protest with armed insurrection. At the same time, these unionists were exposed to the more moderate social democratic ideas coming from their European labor allies, who had provided them with sustained financial support during the war.Footnote ³⁷

Democratization and the end of the war presented left-oriented labor unions with new challenges and raised questions about how unionists understood the world and their role in it. A prominent former guerrilla commander, Joaquín Villalobos, claimed that labor unionists should abandon their classist ideology and learn to cooperate with capital.Footnote ³⁸ Likewise, Commander Fermán Cienfuegos (Eduardo Sancho), reflecting a similar shift in thinking in one segment of the guerrilla Farabundo Martí National Liberation Front (FMLN), told his union followers after the war that unions were too confrontational and activists should think of new ways to organize workers.Footnote ³⁹ Other former guerrilla leaders, while maintaining their traditional classist views on unionism, were so preoccupied with their new tasks as parliamentarians that they had little or no time to dedicate to the union movement. One left-oriented union leader explained that while unionists ideologically were “still plugged into the party, the electricity had been cut off.”Footnote ⁴⁰

In the aftermath, some union organizations, like FENASTRAS, abandoned their leftist orientation and began collaborating with employers. Others held to the belief in radical social transformation, albeit now through nonviolent means, while also embracing autonomy from the political left. For example, the Federation of Associations and Independent Unions of El Salvador (FEASIES) refers to itself as “anti-neoliberal, autonomous from political parties, and a promoter of values such as fraternity, solidarity, and gender equality on local, national, regional, and international levels.”Footnote ⁴¹

The challenge for these left-oriented unionists was to use their understandings and beliefs to interpret the challenges presented by industrial restructuring and state transformation and to develop new union actions. The lead organizer of several campaigns for union formation in EPZs concluded that it would not be possible for workers to organize unions in the EPZ sector “without first having accumulated a good correlation of national and international forces to their favor.”Footnote ⁴² He also insisted that successful activism entailed alliances with nonlabor civil-society organizations.Footnote ⁴³

These union activists had learned during the war that international solidarity could put pressure on governments that supported the war, most notably the United States. Activists also learned during the war that alliances—such as with mothers' committees of the disappeared—provided legitimacy to their efforts while also effectively shaming the military regime. In the new context, these labor activists concluded, instead of pressuring the US government, American activists could pressure American multinational corporations, while women's groups and others could provide added legitimacy and organizational experience to the campaigns. Moreover, unionists, at times unwittingly, developed a more gendered perspective of their work as a result of their contact with women's groups, which was important given the feminization of the workforce.

Other Salvadoran civil society organizations attempted to address the problems facing workers in EPZs. During the period of heightened violence, several human-rights groups were formed to denounce the disappearances, imprisonment and torture of activists, and targeted assassinations. By the 1990s, as the violence subsided, groups such as the Human Rights Office of the Jesuit University and the Catholic Archdiocese began documenting violations in EPZs and providing legal assistance to workers who were illegally dismissed or were otherwise victims of abusive treatment by management.

This panoply of left-oriented activists gathered in the summer of 1995 at the office of the Center for Labor Studies (CENTRA) in a middle-class neighborhood of San Salvador to discuss how to work together to address the labor, gender, and human-rights concerns in the country's booming apparel export sector. They concluded that there was a need for greater coordination among a range of social organizations to target both local employers and the state, and that there needed to be international pressure placed on the multinational corporations that provided the production orders to local factories. If the factories refused to allow workers to unionize, and if the government failed to enforce its own laws, then perhaps international organizations could embarrass brand-name corporations by linking them with the range of violations taking place in the EPZs. To facilitate these efforts, they formed the Coordinating Body for Dignity in the Maquila (COSDEMA).

A crucial international ally of COSDEMA was the New York-based National Labor Committee (NLC), a five-person, $250,000-a-year, activist organization.Footnote ⁴⁴ The NLC was formed in 1980 by a group of New York unionists who were concerned both about US support for the military regime in El Salvador, and American Federation of Labor-Congress of Industrial Organizations (AFL-CIO) support of that foreign policy. According to the Committee's first director, David Dyson, “Almost all of us involved with the local committee and eventually the national committee had been in some way affected and traumatized by the war in Vietnam.”Footnote ⁴⁵ Dyson was greatly disturbed by then-AFL-CIO president, George Meany's, unflinching support of US war efforts in Vietnam, and he did not want to see the entire labor movement go in that same direction with Central America. The NLC organized fact-finding delegations to El Salvador and wrote several reports which highlighted the dramatic state of affairs for unionists in the country, including unjust imprisonment and death-squad assassinations of union leaders.

Charlie Kernaghan replaced Dyson as director in 1989. As the war drew to a close, the disappearances, arrests, and assassinations of labor activists were no longer a significant issue, and thus ceased to be the core concern of the NLC. As news of labor-rights violations in EPZs spread, Kernaghan decided that, rather than terminate the work of the NLC in Central America, the better strategy would be to build on existing contacts and activist networks and shift the group's focus to labor conditions of workers making apparel for American corporations.

What Kernaghan could offer the Central Americans was his ability to put pressure on American corporations through media exposés that framed the issues in ways that resonated with US societal norms, such as respect for human dignity. In his words, “You can have the greatest issue in the world, but if you don't know how to move it, that issue doesn't exist. […] We do thorough research. But then we translate it into a human story.”Footnote ⁴⁶ Kernaghan also realized that one way to get the human story across was to bring young workers to the United States on speaking tours. He explained, “We had faith that these kids would simply tell the truth, and that would be more damaging than anything an academic could say.”Footnote ⁴⁷ Kernaghan traveled to El Salvador several times a year and held lengthy meetings with unionists and other activists to plan strategies. With Kernaghan was his collaborator, Barbara Briggs, who had lived in El Salvador and worked with labor unions there. She was a familiar face to the Salvadoran activists, was fluent in Spanish, and provided a needed connection to the mostly-female workforce in the EPZs.

The National Labor Committee and COSDEMAFootnote ⁴⁸ combined local activism with international pressure. Several of the most notable campaigns targeted the Mandarin, Gabo, Do All, and Tainan factories by putting pressure on MNCs like The Gap, Liz Claiborne, and Nike. In my survey of labor actions in the EPZ sector between 1990 and 2002, I found that seventy-one percent of campaigns to form labor unions combined local organizing with some degree of international pressure on MNCs.Footnote ⁴⁹

The form of this activism was different from traditional union organizing in many regards. First, the main point of leverage was not the ability of workers to disrupt the production process—although that was important too—but rather, the ability of activists to tarnish (or threaten to tarnish) the image of MNCs. Alliances with local nonlabor civil-society organizations added needed legitimacy to the activism. When a religious group or a human rights organization makes claims that workers have been physically abused or sexually harassed, it is less likely that opponents can argue that activists are pursuing their economic self-interest, as they might do if the unionists attempted to pursue the action alone. The other important feature of this new activism was that it did not take the form of a formal organization, but rather of a loose network that shared information and collaborated on campaigns.

Many of these campaigns were relatively short-lived, and their impact was limited to the successful formation of several unions and a reduction of the more substantial rights violations. In addition to the tremendous mobility of the sector, the high burnout rate among the young workforce resulted in a very high level of job rotation. Many workers remained in the sector for only five or six years. As a result, it was very difficult to accumulate experiences and strengthen unions. Most of the workforce had limited formal education and no foreign-language skills, which hampered their ability to develop sustained and informed relationships with international solidarity groups.

While the EPZ sector was fast becoming the largest manufacturing sector in the country, not all unionists were equally preoccupied with organizing EPZ workers. The General Confederation of Trade Unions (CGS), which was formed in 1958 under the military government of Colonel José María Lemus, had been considered the “official” labor center in the country. In a clear rejection of the more confrontational tactics of the left, it pursued its goals through collaboration with employers and the government. Anibal Somoza led the CGS. The walls of his office are adorned with military photos. One shows a group of soldiers proudly displaying their automatic rifles as they stand in front of an attack helicopter. Anibal Somoza's brother was a high-ranking officer in the Salvadoran military. Sitting down in front of his desk reveals another choice of office adornments. Munitions are neatly lined up from large to small along the very outer edge of his desk (which made note-taking during the interview rather awkward for fear that a wrong move could bring the collection tumbling down). Somoza explained that the CGS was not interested in organizing the EPZ sector, which was primarily female. Rather, it preferred to focus its attention on the (male dominated) construction sector.

One conservative union, the Industry Union of Cotton, Synthetics, and Related Industries (STIASSYC), did set its sights on the apparel export sector. Between 1993 and 1996, it organized four unions in EPZs.Footnote ⁵⁰ However, the first two unions—in S&C Apparel and Mandarin International—were later destroyed after management fired all union members including the entire union leadership.Footnote ⁵¹ With the third attempt, in Autran Zacarias, management pressured the union activists to accept their severance pay and leave their jobs. In one case, a women's group approached the union and offered to assist it, yet the union did not trust the organization out of fear that it would take away their members.Footnote ⁵² While there was some interest in working with international activist organizations like the NLC, the union realized that its conservative past did not make it an organization with which the NLC would want to work.Footnote ⁵³

A different organization took another approach. When foreign unionists cut off their funding due to concerns of corruption, FENASTRAS's leaders sought financial rewards by building ties with local management. They did so by relying on the threat of unionization created by leftist unions. Management believed that by letting in a corrupt union that they could control, they could avoid unionization by more radical unions. In 1995, FENASTRAS entered into a pact with Taiwanese and Korean EPZ factory owners in which it offered “industrial peace” in exchange for monthly payments. For a sum of money that depended on the size of the factory, FENASTRAS agreed not to organize strikes and to keep radical unionists out of the factories. FENASTRAS also vigorously denounced the international campaigns by the progressive unionists and their allies, proclaiming that these campaigns had resulted in the loss of jobs and were part of a protectionist conspiracy on the part of US unions.Footnote ⁵⁴ This rhetoric gained the organization the favor of the employers and the government.

To summarize, Salvadoran labor unions facing industrial restructuring and a highly adverse political and institutional context were forced to dramatically rethink their strategies. Left-oriented unionists found that the most effective labor strategy combined base-level activism with international pressure on apparel transnational corporations (TNCs). They learned that they needed to target the nodes of corporate control depicted in Gereffi's buyer-driven commodity chains. This was facilitated by framing labor issues as violations of fundamental rights. Alliances with reputable civil society organizations such as women's groups and human rights organizations provided added legitimacy to the campaigns. Yet not all unionists responded in the same manner. Some conservative unionists decided that, instead of developing new strategies to organize the mostly female workforce in the growing EPZ sector, they would remain focused on traditional male-dominated sectors of the economy like the construction sector. Other conservative unionists did attempt to organize EPZ workers, but they did so without international alliances. This shunning of internationalism was partly based on a belief that the economic interests of activists in developed countries were naturally at odds with the interests of workers in developing countries. It was also a result of historically-rooted political distrust that lingered from the Cold War era. These conservative unionists found that they could organize members by presenting themselves to factory owners as a moderate alternative to the left-oriented unionists. These varying Salvadoran labor responses to globalization and declining state protection are rooted in union histories and lingering ideological differences.

Brazil

Although facing similar economic and political challenges, there are striking differences between the Brazilian and Salvadoran labor movements, suggesting that labor responses to restructuring would take on unique forms in Brazil. Conducting interviews with the Brazilian unionists reveals some of these differences. Walter Sanches, a leader of the country's largest metalworkers' union and the head of the DaimlerChrysler unit, greets researchers by politely inquiring, “What language would you like to do this interview in, Portuguese, English, or German?”Footnote ⁵⁵ Following the interview, one is guided to a research department with countless documents on the industry and labor activism that go back at least thirty-five years. Full-time economists and lawyers assist the union in designing strategies and preparing for negotiations.

The Brazilian unionists in the auto producer-driven commodity chain had been doing this work for a very long time. They traveled frequently to international union events, and several of the top leaders had decided years before that they needed to learn the language of the multinational auto company for which they worked. This would not only help them to negotiate with international corporate managers, it would also facilitate sustained international labor solidarity. The level of sophistication, organization, and skill displayed by the Brazilian auto unionists are the product of years of evolution and a strong political and ideological commitment to their work. These qualities are also a byproduct of the auto industry itself, which has deep roots in Brazil and has provided stable careers to a generation of workers. Yet, like the Salvadoran apparel-export workers, industrial restructuring presented the Brazilian unionists with dramatic new challenges and the need to explore new labor actions.

While the first automobile plants were installed in Brazil in the 1920s, the industry achieved national significance in the 1950s when then President Juscelino Kubitschek (1956–61) united political and economic groupings in the country around the concept of “national developmentalism.” Automobile production was envisioned as the pathway to modernity, and the government used every available policy instrument to pursue the expansion of the industry, including hefty domestic content requirements and a ban on car imports. By 1961, Ford, General Motors, Volkswagen, and Mercedes Benz were all producing in Brazil. Military rulers further encouraged the expansion of the auto industry, and the sector soared. Then the oil crises of the 1970s were followed by the debt crises of the 1980s, and the industry stagnated.

The return to civilian rule in the 1980s was followed by a period of government policies designed to increase Brazil's ties to the global economy, in part by internationalizing the automobile industry. The average tariff on imported autos dropped from 85 percent in 1990 to 34.3 percent in 1994, and it continued to drop throughout the 1990s.Footnote ⁵⁶ As a result, the number of imported cars jumped from 6,100 in 1990 to 420,000 in 1995.Footnote ⁵⁷ The pressure on auto manufacturers to compete resulted in the use of outsourcing to reduce costs and increase competitiveness. Manufacturers also increasingly relied on multinational suppliers for parts and components. And all major plants built in the late 1990s and early 2000s were built outside of the core industrial district of greater São Paulo in regions where unions were weaker and wages were lower.

At a modular Volkswagen production facility in Resende (outside Rio de Janeiro), there is not a single Volkswagen production worker in the plant. A network of independent contractors provides all the components and performs the auto assembly. A small group of Volkswagen employees are on hand to inspect the vehicles during the production process. The company concentrates on design, engineering, and marketing.Footnote ⁵⁸ According to Elder Couto, a manager at the Volkswagen plant in Rio, “building the product is not our core competency.”Footnote ⁵⁹

The Ford Motor Company, located its most modern facility in Brazil in the northeastern state of Bahia, where twenty-seven suppliers are located inside the facility and directly feed the assembly line.Footnote ⁶⁰ Some sixty percent of the 3,297 employees inside the facility work for suppliers. Between 2000 and 2002, workers in new modular production facilities on average earned $445.79 per month compared to $807.92 per month earned by workers in traditional auto-production facilities. Unionization rates were likewise much lower. Some 65.85 percent of workers in traditional auto plants were union members compared to 34.79 percent in modular plants.Footnote ⁶¹

And in the sector's stronghold, the ABC Region outside of São Paulo, outsourcing and segmentation have led to an inability to defend jobs and regular wage increases through traditional union activities such as strikes. Between 1980 and 1999, the combined employment at Ford, Volkswagen, and DaimlerChrysler in the region dropped from 66,541 to 31,850. And from 1994 to 2002, salaries dropped from $1,023 per month to $876 per month.

Thus, while the Brazilian union movement was more developed than the Salvadoran movement, it faced many of the same problems of wage and union decline due to industrial restructuring. And like in El Salvador, labor's ability to respond to these new challenges was partially shaped by the country's labor institutions and practices. Brazil developed a system of corporatism in which the state played a central role. For John Humphrey, the goal was to “ensure social harmony by guaranteeing the interests of both labor and capital under the tutelage of the State.”Footnote ⁶² In the words of President Getúlio Vargas, during whose authoritarian rule (1930–1945) Brazilian corporatism took shape: “The State does not want and does not recognize class struggle. The labor laws are laws of social harmony.”Footnote ⁶³ Brazilian corporatism combined monopoly representation with decentralization. Unions were formed at the subnational (mostly municipal) level.Footnote ⁶⁴ As John Humphrey explains, the system changed unions from independent defenders of workers' interests to “financially secure, bureaucratic welfare agencies.”Footnote ⁶⁵ These moderate, state-dependent unionists were known as pelegos.Footnote ⁶⁶

When the military took power in 1964, it declared all strikes illegal, placed many unions under the control of the Ministry of Labor, and arrested leftist union leaders. The context for labor shifted in the mid-1980s as the country democratized. The new civilian president, José Sarney, relaxed control over strikes and in 1988 enacted constitutional reforms that were considered favorable to labor. The state no longer registered unions or intervened in unions' internal affairs, public-sector workers could organize and strike, a profit-sharing mechanism was introduced, and workers in firms with more than 200 employees were allowed to form a workers' council to negotiate with employers.Footnote ⁶⁷

Yet the reforms left numerous components of the corporatist regime intact. The labor code controlled many aspects of union activities, and the labor courts held broad powers over the terms of collective bargaining and strike activity. For Maria Hermínia Tavares de Almeida, “the 1988 Constitution […] contributed to create a very strange hierarchical structure combining corporatism at union level and pluralism at the top.”Footnote ⁶⁸ Brazilian labor laws continued to favor union formation at the municipal level, so that as industry expanded nationally unions were not able to easily organize new plants or automatically extend collective bargaining agreements to new work centers. This was particularly challenging for auto unionists as the sector became dispersed throughout the national territory in the 1990s.

Even when the law was favorable, employers used other illegal means to avoid union representation. For example, some companies compiled blacklists by gathering information from labor courts on workers who filed lawsuits against companies for any reason. The practice served to discriminate against workers who used their constitutional rights to present a complaint, while also scaring other workers from using the court system should they have a claim against an employer. The situation was even more severe outside the large urban centers. The International Labour Organizations' 2005 Committee of Experts report noted, “local inspection offices […] are ill-equipped and inadequate and often lack computers, telephones, and even the most basic furniture (tables, chairs) needed by inspectors to carry out their duties.”Footnote ⁶⁹ The report also noted that in January 2004 three labor inspectors and a driver were killed while attempting to carry out their duties.Footnote ⁷⁰

In this context of outsourcing, segmentation, and inadequate state protection, Brazilian auto unionists pursued new strategies. These strategies were informed as much by economic and political structures as by historical experiences and ideational influences. In the early 1900s, unionists were motivated by anarcho-syndicalism as the largely immigrant working population—who did not enjoy voting rightsFootnote ⁷¹—favored factory struggles and strike activity over electoral politics. Yet, following the Russian Revolution and as anarcho-syndicalists' direct action was increasingly unable to resolve workers' demands, left-oriented workers turned to communism. In 1929, they formed the General Confederation of Brazilian Workers. Like their comrades in El Salvador, the Brazilian Communists during this period were committed to the class-conflict thesis, and in 1935 they participated in an armed uprising that was put down by the regime.

When Getúlio Vargas became president of Brazil in the 1930s, the state encouraged the formation of moderate unions, which displaced the anarchist and communist unions that had dominated the movement until that time. Decades later, the military's policy of promoting concentrated industrialization in the São Paulo region created conditions for the leftist movement's reemergence. As John Humphrey explains, by 1970 in São Bernardo (the industrial suburb of São Paulo), “73 percent of all workers in manufacturing industry were represented by one union, and over half of the workers in it worked in the five large auto plants.”Footnote ⁷² By the late 1970s, as the military loosened some of its controls on society, labor unions took advantage of limited liberalization to mobilize and demand full democratization.

The movement did not pursue the radical transformation of society through socialist revolution.Footnote ⁷³ Its demands were shaped by their rejection of state corporatist controls while their actions were influenced by new social activism motivated by Christian Base Communities.Footnote ⁷⁴ During the unrest of the late 1970s and early 1980s, these workers developed alliances with numerous civil-society organizations, including the progressive wing of the Catholic Church, while remaining independent of the state and political parties. They formed what became known as “new unionism,” which was institutionalized in the left-oriented labor central, the Unified Workers Central (CUT).

With time, the movement developed an oppositional or class-based ideology. According to its 2002 bylaws, the CUT “is a labor union organization of the masses of the maximum level that is classist, autonomous and democratic.”Footnote ⁷⁵ Within its general adherence to a left-oriented ideology, there were numerous ideological trends within the CUT that ranged from radical Trotskyist groups to moderate social-democratic tendencies. Over the course of the 1980s and 1990s, the CUT shifted its tactics from a more confrontational posture to a more conciliatory position. Strike activity declined after labor leaders such as Luiz Inácio “Lula” da Silva joined with other progressive activists to form the Worker's Party (PT) in 1979. Through the PT, union leaders ran for and won political office. In 2002, Lula became the first blue-collar worker and union leader to be elected president of Brazil. During this era, the CUT grew and became more institutionalized. It maintained its general criticism of the capitalist status quo, but adopted more moderate means to pursue its goals of social transformation.

In the Brazilian auto sector in the 1990s, the left-oriented unionists of the CUT concluded that the most effective strategic response to industrial restructuring was greater international coordination. The biggest challenge facing the union was job loss in the highly-unionized plants outside of São Paulo. In theory, job loss is one of the most difficult issues around which to develop international cooperation, due to its zero-sum nature (one group's job loss is potentially another group's job gain). Yet in the German automobile plants in Brazil, unionists had been developing a strong relationship for over two decades with their union colleagues in Germany, which was facilitated by the language abilities of top Brazilian unionists. It was the strength of this relationship that allowed them to navigate a solution to the challenges of industrial restructuring.

Initial contacts between Brazilian and German unionists working for Volkswagen were established in 1978 at an international union conference. It was here that the German unionists learned about the repression faced by Brazilian Volkswagen workers under the military dictatorship.Footnote ⁷⁶ Four years later, when Volkswagen attempted to establish a works-council system in which only fifty percent of the members could be unionists, the Brazilian unionists contacted their German colleagues whom they had met at the conference. The Germans agreed to visit the country. When they returned to Germany, they demanded that the company accept “a legitimate system of representation in Brazil.”Footnote ⁷⁷ A works-council structure was subsequently established in Brazil that did not restrict membership from the union.

The relationship between the unionists was deepened in 1984 when a delegation of Brazilian unionists from Volkswagen spent seventeen days in Germany meeting with German union and works-council representatives.Footnote ⁷⁸ The trip to Germany was followed by trips by German unionists to Brazil in 1986 and 1987. With the support of German unionists, the Brazilian Volkswagen workers achieved the right to strategic technical information, which was a breakthrough for the union because it allowed for more effective collective bargaining.

Labor relations at the Volkswagen plant in São Bernardo were relatively harmonious for much of the 1990s. However, the situation changed following the Asian financial crisis. Responding to the downturn in sales, Volkswagen proposed to the union the implementation of a shorter and more flexible workweek (known as the “Volkswagen Week”). The union agreed to the proposal, but on a temporary basis. In 2001, as the company's financial concerns continued, it sought to extend the Volkswagen Week. The union rejected this proposal, arguing that the crisis did not appear serious enough to merit such an action.

In response, Volkswagen announced that it would dismiss 3,075 workers at its plant in São Bernardo. Soon afterwards, the 16,000 workers of the plant voted to go on strike. Yet the union leaders realized that this traditional strategy would not resolve the problem. Indeed, a strike might even be welcomed by the company as an opportunity to sell off some of its inventory while not having to pay production workers. The challenge was to get new production orders into the plant to protect jobs, and the decision concerning where to produce new car models was made in Germany. The union noted, “Since Volks [management in Brazil] insisted on the firings and gave no sign of wanting to reopen the negotiations, we made the decision to go to Volks headquarters in Germany.”Footnote ⁷⁹ That is, the auto unionists realized that they needed to directly engage the production headquarters in the producer-driven commodity chain.

After several days of negotiations in Germany between the German management and the Brazilian union leaders, Luiz Marinho and Mario Barbosa (who speaks fluent German), the two sides reached an agreement. Some workers would receive paid leave and might be allowed to return to work later. Other workers would be let go for underperformance, and 1,500 workers would go back to work immediately.Footnote ⁸⁰ This was the first time in forty-three years that an agreement was made in Germany. German unionists played an active role in assisting the Brazilians during the negotiations, and thus helped the Brazilian unionists defend their jobs. This level of international union coordination did not happen quickly. It was the product of more than twenty years of communication, trust-building, and solidarity actions around less potentially conflictual issues, such as protesting military intervention in labor conflicts and defending the right of workers to form autonomous factory commissions.

Similar sustained alliances between unionists at Mercedes Benz (later DaimlerChrysler) in Germany and Brazil also allowed for coordination to avoid potential job loss while strengthening mechanisms of worker representation. At Ford, the relationship with American unionists was not as strong due to the United Autoworkers' (UAW) more limited engagement with sustained labor internationalism.Footnote ⁸¹ Yet, in 2001, with the assistance of the UAW, Brazilian union leaders traveled to Detroit in order to leverage the production headquarters of Ford's producer-driven commodity chain. They met directly with Jacques Nasser, the CEO of Ford, and negotiated a five-year guarantee of employment for Brazilian workers in the São Bernardo plant.

A conservative union, the Metalworkers' Union of São Paulo (SMSP) took a different approach to the challenges presented by industrial restructuring in the auto sector. This union grew out of Brazil's corporatist past. It was a founding member of Força Sindical, what one observer refers to as a “conservative modernizing” union tendency.Footnote ⁸² Força Sindical and the SMSP have been critical of left-oriented unionism and international union solidarity. Rather, their response to industrial restructuring has been to seek deeper ties, when possible, with employers. Indeed, in the mid-1990s, Força Sindical is reported to have received approximately $5 million from corporations such as Alcoa, Shell, Siemens, White Martins, and Souza Cruz in exchange for its management and market-friendly policies. Força Sindical May-Day events have drawn hundreds of thousands of participants in part because they raffle off cars donated by major auto manufacturers. In general, Força Sindical shunned confrontational actions and international labor alliances, preferring pacts with employers and state entities.

The leader of the Força Sindical's metalworkers' union, Ramiro de Jesus Pinto, is open about the union's strategies. He proudly outlines the union's many social programs, including a housing project for workers. Unlike the metalworkers in the left-oriented CUT union, the SMSP has no regular contact with unionists in Germany, the United States, or Canada, nor much less has it coordinated international strategies to prevent plant closings or to strengthen local works-council structures. When asked about the International Metalworkers' Federation, the parent organization for auto unions (to which SMSP is affiliated), the president asserted that it was worthless. Ramiro held strongly to the belief that international labor solidarity in a globalized economy was “impossible due to competition over jobs.”Footnote ⁸³

One important case for SMSP involved the Ford Ipiranga plant in São Paulo. The Ipiranga plant, which began functioning in 1953, was Ford's oldest plant operating at the time. By 1999, in the aftermath of the Asian financial crisis, the Brazilian auto sector faced a dramatic drop in sales. Ford began cutting jobs in São Paulo and São Bernardo. Despite a series of activities and the union's attempt to negotiate,Footnote ⁸⁴ the plant shut down. According to the representative of Força Sindical, “there was nothing left to do. We did not seek to develop alliances and pressure Ford in the U.S. There were no international negotiations. We just negotiated the closing.”Footnote ⁸⁵ At the General Motors factory in Gravataí, the Força Sindical union has been accused of collaborating with management to keep the left-oriented union out. It used the management's fear of unionization by the CUT unions to strike a deal through which it was able to gain representation of the workforce.Footnote ⁸⁶

To summarize, the combination of market-oriented reforms and inadequate labor laws in Brazil created new challenges for auto unionists. For left-oriented unionists, the militant strike activity that had worked so well during the dictatorship was less effective in a globalized era in which the auto industry faced a crisis of overproduction. Left-oriented unionists learned that, rather than fall into an international competition over jobs, they could build international labor alliances with unionists in other countries to ensure that industrial restructuring was the least-damaging possibility to all workers concerned. In some cases, they were able to pursue direct negotiations with the production headquarters of a global auto commodity chain with the assistance of the unionists in those countries. This level of internationalism surpassed that of the EPZ unionists in El Salvador. It was the product of over two decades of unionists across countries working together to slowly develop the trust and shared understandings that would be necessary to forge a common international union strategy against job loss.

For conservative auto unionists in Brazil, market-oriented reforms and state transformation made national-level corporatist pacts between labor and the state less tenable. These unionists faced the same problem of job loss that the left-oriented unionists faced. Yet, they did not believe that unionists in other countries could help them solve their problems. At times, corporations provided funds for union activities, including large social events that the conservative unions used to attract members. And, like some of the conservative unionists in the Salvadoran EPZ sector, conservative auto unionists in Brazil were able to use the threat of unionization by left-oriented unionists to gain the favor of employers and unionize new factories. The divergent responses of Brazilian unionists in the automobile industry to the dramatic restructuring of the sector is a reflection of past experiences, differing labor ideologies, and sustained personal relationships in Brazil and abroad that had been established over decades.