Internet: lfrench@vt.edu

Disputes regarding the ethics of child labor have intensified in recent years as publicity about some of the abusive child labor practices has provoked moral outrage around the world. Moreover, growing exports from developing areas have brought charges by labor groups in the United States and other importing countries of unfair competition based on extensive employment of children. Unfortunately, more heat than light has been generated because the discussions fail to recognize the diverse forms that child labor takes or to be based on systematic empirical research. After reviewing central issues in these debates, we focus on child labor in export industries and report on data gathered in Brazil's export-oriented shoe industry that has, in the past, utilized the labor of many children and adolescents. Central findings are: 1) the main causes of child labor at this site have less to do with underdevelopment and more to do with how the shoe industry and its workers are integrated in the global order, 2) the form that child labor takes in this industry makes it appear rather benign: children want to work, frequently are employed by their parents, and remain in school, but 3) both children and adult workers suffer from the employment of underage labor.

Disputes regarding the ethics of child labor have intensified in recent years as publicity about some of the abusive child labor practices has provoked moral outrage around the world. Moreover, growing exports from developing areas have brought charges by labor groups in the United States and other importing countries of unfair competition based on extensive employment of children. Unfortunately, more heat than light has been generated because the discussions fail to recognize the diverse forms that child labor takes or to be based on systematic empirical research. After reviewing central issues in these debates, we focus on child labor in export industries and report on data gathered in Brazil's export-oriented shoe industry that has, in the past, utilized the labor of many children and adolescents. Central findings are: 1) the main causes of child labor at this site have less to do with underdevelopment and more to do with how the shoe industry and its workers are integrated in the global order, 2) the form that child labor takes in this industry makes it appear rather benign: children want to work, frequently are employed by their parents, and remain in school, but 3) both children and adult workers suffer from the employment of underage labor.

keywords: child, labor, Brazil

In recent years the topic of child labor has emerged as a leading issue in the field of international business ethics. Abuse of children in the workplace can trigger strong emotional reactions and concern about this practice has energized people to protest corporate practices who would not normally consider themselves social activists. This is, of course, understandable as it is difficult not to be struck by the irony of individuals in developed countries purchasing such products as athletic shoes and apparel, soccer balls, and other sports equipment made by impoverished children in the Third World for their own consumption or that of their relatively affluent children.

When one gets beyond the initial visceral reaction against child labor and begins to explore the phenomenon more carefully, it becomes clear that there are many complex and contentious issues associated with it. This complexity has led some commentators to suggest that those who criticize child labor practices in the Third World are at best naïve, and at worst are engaging in hypocrisy, protectionism, and a sinister form of ethical imperialism.

To move forward the discussions, it is first necessary to recognize the diversity of forms which work by children takes. The International Labour Organization (ILO) estimates that about 73 million 10-to-14 year olds work throughout the world (ILO, 1996). Additional millions of even younger children are believed to work as well, but accurate estimates are even harder to make. Most children who work do so for their own families in agriculture or domestic activities (UNICEF, 1998). Significant numbers are also found in small, labor-intensive manufacturing firms, in mining activities, and in plantation agriculture. Still others, relatively few in number, but widely discussed, labor as child prostitutes or survive on the street through a variety of illegal or illegal means. Finally, some fraction of child laborers, perhaps 5% of the total, work in activities linked to global industries.

In this paper, we begin by discussing some of the issues dealing with definitions of child labor and then provide an overview of the ethical debates over this phenomenon to date. We then focus on the widely-publicized issue of child labor in emerging global issues. Studies suggest extensive employment of Third World children in consumer goods industries organized to produce for export to the markets of already-industrialized nations (US Department of Labor, 1994). Such work raises a host of questions regarding appropriate definitions of child labor, challenges conventional notions of its causes and consequences, and brings to the fore the issue foreign stakeholders' roles. To ground our discussion, we look at the issue in the shoe industry of Brazil where problems of child labor have received much attention and a program designed to combat child labor has received favorable reviews. We will consider the impact of this program, provide an ethical analysis of child labor in the town where it operates and offer some suggestions for differentiating between morally acceptable and morally unacceptable child labor in this particular setting.

Previous fieldwork undertaken by the authors has focussed on the causes and consequences of child labor in the shoe industry of Brazil. The approach taken to this analysis presented here is consistent with the grounded theory perspective of research (Glaser and Strauss, 1967; Strauss and Corbin, 1990). This is the idea that theory development should be guided by empirical observation. This research approach is well accepted and widely followed, but is noticeably lacking in most ethical analyses of child labor. To date, there has been surprisingly little systematic, empirical research that examines why children work (White, 1994; Weiner, 1991). By combining and building upon previous work conducted by us we believe that we will be able to make a contribution to the understanding of the phenomenon of child labor.

Discussions of child labor are hampered by imprecise and conflicting definitions of what constitutes child labor. Confusion is focussed primarily on two issues: 1) What is the age criterion for differentiating between an adult and a child? and 2) What types of work by children should be considered abusive and undesirable and what types should be considered benign or even beneficial?

In everyday discussion, the term "child" has different meanings for different people, cultures, and historical periods. However, several international conventions have helped provide standards for clarifying confusion over this issue. First is the ILO Minimum Age Convention, 1973 (No. 138) that states, "The minimum age…shall not be less than the age of completion of compulsory schooling, and in any case, shall not be less than 15 years." Second is the UN Convention on the Rights of the Child, 1989, which specifies that, a child means "Every human being below the age of eighteen years unless, under the law applicable to the child, majority is attained earlier."

The second point of confusion, regarding the nature of the work, is more problematic. Perhaps the broadest notion would include virtually all paid work activities undertaken for corporate employers by those under a certain age. In practice, this has been modified to allow some limited number of hours of such activity. Laws of many countries seek to regulate the work of children in part by specifying the maximum number of work hours permitted and, initially, empirical research of child labor focused on the relationship between hours worked and outcome variables of interest (e.g. children's health or grades in school). In recent years, critics of this approach have argued that attention should be directed less to the mere fact of work or the hours of work and more to the nature of the work activities themselves. This argument has led to the idea that child labor involves work that threatens children's development while child "work" constitutes activities that foster their skill development and socialization to modern work values and attitudes. Senator Tom Harkin makes this distinction in arguments he made in support of the Child Labor Deterrence Act he is sponsoring in the US Congress:

This is a useful differentiation in that it focuses on the nature of the work (i.e., exploitive or hazardous work) and the results of the work (i.e., harm to children). However, Harkin’s examples only serve to demonstrate the difficulty of operationalizing this distinction. Certainly mining and industry (presumably, factory work) are not the only potentially hazardous sectors of the economy. In fact, farming is quite hazardous. And one can even envision hazardous or oppressive conditions imposed on children selling newspapers or shining shoes. The ILO also asserts the importance of distinguishing between safe and hazardous child work. However, this organization also notes that in practice this distinction is difficult to make.

A related problem regarding definitions of child labor is the treatment of "family work." To date, virtually all definitions of child labor exclude work activities undertaken within the family. However, UNICEF (1998) suggests that much of the work done throughout the world by young people occurs within family enterprises. Jobs in these businesses are typically regarded as opportunities for children to develop useful skills and work attitudes under the watchful eyes of protective parents. Assumptions that the children are unpaid reinforce the notion that this family work is somehow unrelated to the labor market even though the family business operates in the market economy. Thus, by definition, family jobs constitute child "work" and not "labor."

Even more worrisome is the failure to consider the possibility that performance of domestic chores might threaten the development of young people and therefore fall into the category of "labor" rather than work. Many argue that children in developing areas, especially girls, are forced to spend such time on household chores that their ability to learn in school is seriously compromised. Although rarely paid, these activities are clearly related to the labor market in the sense that they release adults, typically the mother, for outside work.

Underlying the lack of attention to family work, either jobs in enterprises or chores in the house, is the assumption that families function as cooperative systems in which all parties have similar interests and in which parents act in the best interests of their children. The possibilities of conflict between family members and exploitation of the weaker by the stronger are, of course, unrecognized by those operating under this assumption (Nieuwenhuys, 1996).

In recent years much of the criticism of child labor practices in developing countries (as well as a host of other labor practices said to violate human rights) has come from social activists and labor organizers in the United States and Western Europe. These countries have much higher per capita income and different cultural patterns than exist in the developing countries and the critics are sometimes accused of not taking these differences into account in their criticisms. Defenders of child labor practices in the Third World point to certain hypocrisy in these criticisms, noting that these developed countries also used child labor extensively when they were at a similarly early stage of development. Indeed some of them still have a high rate of labor force participation by children. Noting the participation of organized labor in efforts to outlaw the importation of goods made with any child labor these commentators have further argued that these efforts are motivated more by protectionist sentiments than by humanitarian concerns. (See for example, Rockwell, 1999; Stelzer, 1999; and Weidenbaum, 1999 for variations on these arguments.) And finally these critics are also accused of engaging in ethical imperialism, that is, trying to impose their ethical standards on those who live in other countries without regard to local conditions and practices.

Thomas Donaldson, in his Harvard Business Review article, "Values in Tension: Ethics Away from Home" (1996), provides a framework for evaluating these charges of ethical imperialism (if not the critics’ motives) and more generally for analyzing ethical dilemmas regarding international business. He notes that differences in business practices between home and host country are quite common for multinational corporations (MNCs), especially when the company’s home country has a much higher economic standard of living. He further notes that it is important for MNCs operating in such diverse environments to avoid ethical imperialism as well the opposite extreme of ethical relativism (i.e., uncritically accepting actions as ethical just because they are common practice). Donaldson concedes, however, that doing so is frequently quite difficult but he contends that companies will be well served in this endeavor by following three guiding principles: 1) respect for core human values, 2) respect for local traditions, and 3) the belief that context matters.

There are, according to Donaldson, two types of ethical conflicts that can occur when corporations do business in other countries: conflicts of relative development and conflicts of cultural tradition. Conflicts of relative development pertain to situations where the level of economic development of the MNC’s home country is different (in most cases higher) than that of the host country. Conflicts of cultural tradition pertain to differences in culture, mores, and norms that are, for the most part, independent of the level of economic development in a country. Donaldson further suggests different decision-making frameworks are called for in resolving conflicts of relative development versus conflicts of cultural tradition. This makes sense, and suggests that an ethical analysis of child labor should be based in part of an understanding of why children work. Yet this has been noticeably lacking in discussions of child labor. Much of the writing about child labor is editorial in tone, alluding to general ethical principles or broad macroeconomic implications along with anecdotal evidence to support the author’s contentions (see Myerson, 1999; Nader, 1996; Rockwell, 1999; Ryan 1999; USA Today, 1996; and Weidenbaum, 1999).

Arguments in support of child labor have been presented both from a utilitarian perspective and a rights-based ethical perspective. From a utilitarian perspective some have argued that the use of child labor is justified in that it is a necessary step in the progress of these developing countries, like it was when the United States and the developed countries of Western Europe were in similar stages of economic development (see for example, Alam, 1999; Iovine, 1999; Myerson, 1999; Stelzer, 1999; USA Today, 1996; Weidenbaum, 1999, and arguments contained in Nichols, 1993, and Shaw and Hartman, 1999). According to this view, the benefits to society and to the child laborers themselves, particularly in terms of advancement of economic position outweigh any harms of children’s participation in the workforce.

Others argue that those who would seek to end child labor are infringing on the rights of children (and their parents) to do what they want with their lives and to advance their own economic circumstances (Alam, 1999; Nieuwenhuys, 1996; White, 1994). Moreover, they contend that ending child labor will only make matters worse for the child workers who may be forced to work in a more hazardous informal sector of their nation’s economy or engage in illegal activities such as prostitution, drug dealing or petty larcenry.

Opponents of child labor typically object to the practice on the grounds of the harm it does to the child workers. It has been has argued that abusive child labor practices violate both a basic right and an element of human dignity (reference deleted). This is the belief that individuals should be given the opportunity to achieve their potential as human beings and that this opportunity should not be forfeited because of the economic circumstances of their parents. It was further argued that children should not have to work in conditions that foreclose their opportunities for reaching that potential because of lack of access to basic education or because of exposure to hazards likely to endanger the child’s physical well-being.

We concur with this latter argument that abusive child labor practices are a violation of a basic right and an element of human dignity. However, the grounded theory approach to ethical analysis suggested here dictates that we would evaluate the ethics of specific child labor in light of empirical information about the extent to which such abusive conditions (i.e., exposure to hazards, deprivation of access to basic education, etc.) apply. We would also include in our empirical inquiry consideration of the perceptions of the child workers themselves regarding the conditions under which they work. We concede that children, who by definition are not fully rational adults, deserve some special protections, even from their own unwise choices. So it would be misguided to rely exclusively on their views in these matters. Nevertheless, we think it would be equally unguided to ignore their views (Lavalette, 1994).

In sum, we believe that broad-brush assessments of the morality of child labor in general are unfounded. Instead we will make more specific evaluations relative to specific areas for which we have some data on these work (and work-related) conditions. Thus in this paper we examine the phenomenon of child labor and efforts to address problems associated with it in a specific geographic and industrial setting. Before looking at this specific case, we first consider some of the general strategies that have been suggested for dealing with child labor problems.

This view, consistent with utilitarian arguments, is that on balance the situation regarding child labor will take care of itself. According to this position, economic development, spurred in part by relatively cheap child labor, will eventually create the conditions in which child labor is not needed and children will no longer need to work. Higher wage jobs requiring more training and education will be created and children will see that they are better off by going to school than by working (or maybe they will do both).

Supporters of this view concede that there are certain unsavory aspects of child labor, but they contend that efforts to end child labor are misguided and they will only cause more problems than they solve. The example of Bangladesh is frequently cited to support this contention. When the US threatened to cut off the importation of Bangladeshi goods made with child labor, employers in the apparel industry quickly laid off child workers. This of course did not help the child workers and many found themselves considerably worse off, unemployed or with jobs in the informal economy or with illegal work in the drug industry or the sex trades (Alam, 1999; USA Today, 1996).

According to this view, a complete ban on the importation of products made with child labor is necessary in order to discourage producers from exploiting child labor in their production processes. Senator Tom Harkin argues that the United States promotes other values through its trade policies and that failure to afford children the same protections we provide animals and criminals is an unfortunate discrepancy:

In recent years individuals have been very active in using their positions as consumers and stockholders to influence corporations regarding the use of child labor and sweatshop labor. Various groups such as the National Labor Committee, Campaign for Labor Rights, Free the Children, the Child Labor Coalition, the Interfaith Center on Corporate Responsibility, and United Students Against Sweatshops coordinate activities against companies judged to be violating acceptable labor practices. Such groups have targeted a number of companies including Nike, Walt Disney Enterprises, Wal-Mart, and Kathie Lee Gifford for shareholder and/or consumer activism. Boycotts, letter writing campaigns, leafleting, sponsorship of shareholder resolutions, and the divestment of stock of selected companies have all been tactics that have been used by these groups.

The potential of these tactics to influence companies is substantial. It has been estimated that over $2 trillion in investment funds have some type of social responsibility or ethical criteria guiding investment decisions (Social Investment Forum, 2000). And boycotts or the potential for them seem to have been instrumental in producing responses, albeit not always completely satisfying to the protestors, from the companies noted above that have been targeted.

In the case of both consumers and stockholders, access to accurate information on corporate practices with respect to child and sweatshop labor is key. The organizations noted above provide information, sometimes of an anecdotal nature, about corporate labor practices that is used to encourage action against companies perceived to be engaging in unacceptable practices. More systematic information on corporate social performance in general and labor practices in particular is available from organizations like the Council on Economic Priorities (Hertel with Powell and Chowdhury, 1997) and Kinder, Lydenberg, and Domini (2000). Also, the US Department of Labor publishes a list of "trendsetter" firms judged to be making good efforts in addressing sweatshop and child labor problems (Clark, 1999). In addition corporations increasingly have been providing additional information on their own activities in response to criticism (see for example Nike, 2002 and adidas-Salomon, 2001). At the same time though, corporate officials themselves often bemoan the difficulty of getting accurate information on labor conditions when a wide network of primary and secondary contractors makes their products. One method of making information on workplace conditions more widely accessible is through the establishment of industry-wide codes of conduct and the use of product labels as discussed below.

In a 1996 op-ed piece, Ralph Nader (1996) cites evidence that consumers are willing to pay a premium for products that are made without the use of child labor. So the use of labels to certify that products have been made without child labor or sweatshop labor can be a marketing advantage. Increasingly, companies have adopted codes of conduct to govern work activities in their own plants and those of their subcontractors. The prohibition of "sweatshop labor conditions" and child labor are usually components of those codes. The apparel industry that has come under much criticism for allowing these labor abuses to occur has been in the forefront of efforts to promote industry codes and product labels. In the athletic apparel industry Nike and adidas-Salomon both have their own codes of conduct and inspectors who visit company-owned factories and subcontractors to make sure that these codes are followed. They also belong to the Fair Labor Association that is providing third-party verification of code compliance.

From 1970 to the mid-1980s, Brazil's leather shoe industry expanded rapidly to become one of the world’s largest exporters with most of its exports destined for the US market (Fensterseifer, 1995). Subsequently, China's entry into world markets has eroded Brazil's production for both export and domestic consumption. By 2000, both output and employment had fallen by roughly 30%. Nonetheless, 25-30% of the 500 million pairs produced in Brazil continue to be exported and earn nearly US$1 billion per year (Pizzo, 2001).

Efforts to cut costs have led most manufacturers to intensify their subcontracting of labor-intensive cutting and sewing operations to small, family-run shops often started by former employees. Frequently, the owners employ their children and others on a part-time basis. Although children under 16 (raised from 14 in 1999) cannot work legally, government officials have rarely intervened to charge parent-employers with child labor violations. However, since the mid-1990s, the US government has brought pressure to bear on the shoe industry to take steps to eradicate child labor which, it believes, contributes to Brazil's penetration of US shoe markets (US Department of Labor, 1994).

Brazil's shoe industry is organized in a number of industrial districts. One center of men’s leather shoe production is Franca, a city of about 300,000 in the southern state of São Paulo. About 400 shoe manufacturers, 25 tanneries, and 1200 cutting and sewing shops comprise the core of this city's economy and produce about 25 million pairs annually with 25-30% exported to the US. In the mid-nineties, research by the local shoemakers union revealed extensive child labor in Franca (Licursi, 1995). Subsequent surveys by the first author in 1998 and 1999 of young public school students about their after-school work activities substantiate earlier findings (references deleted). These surveys were conducted in two neighborhoods: Vila Aparecida, a working class area containing many shoe firms as well as residences and Jardim Vicente de Paula, a lower middle class area whose residents work in the nearby central commercial district and shopping center. Random samples of classes were taken and all students within each class completed the survey during class time. When the 1998 survey was completed, 73 students with jobs were also interviewed about their work experiences. These interviews were held in empty classrooms during the school day and typically lasted 30-45 minutes. Additionally, from 1997 to the present, numerous interviews have been periodically conducted with a variety of individuals in Franca regarding the issue of child labor. These include school system personnel, local government and union officials, owners and managers of shoe firms, shoe industry association leaders, and directors of programs designed to benefit youth in the city. Table 1 displays the work status of the students surveyed in Vila Aparecida. As this suggests, the incidence of child work in this area is very high.

The poverty perspective (i.e., the notion that children work because they and their families are poor) provides limited help in understanding the causes of child labor in Franca. First, like many other cities in São Paulo state, per capita income is considerably above Brazil's average: US$6000 (see Franca City Hall, 1999) versus $4300 in Brazil as a whole. To some degree, the relative affluence stems from the export success of the shoe industry combined with pressure from the militant union to raise wages. Second, due to intensifying competition for jobs, those children and adolescents who succeed in finding jobs often come from more affluent and well-connected families. Like much of Brazil, significant inequalities in income are apparent in Franca as declining shoe production in the 1990s and government anti-inflation measures have increased unemployment substantially. As a result, many seek jobs in Franca's shoe industry and employers are increasingly selective in their hiring practices. In recent years, they have paid more attention to prior work experience and educational credentials as well as personal relationships they have (or don't have) with the applicants or their parents (Rodrigues, 1997). In this context, financial need may certainly play a role in young people's decisions to seek work, but those successful at securing employment tend to come from somewhat more affluent, working class families whose parents have some resources and relationships with the shoe industry. In the surveys referred to earlier, the respondents were asked about the possession of items in the household as an indicator of their family income. Table 2 compares the family possessions of working and non-working children in Vila Aparecida. This further supports the observation that child workers in Franca are relatively well off.

Industry efforts to increase subcontracting in recent years have led to large numbers of small, family-run cutting and sewing shops or bancas. Founding and operation of such enterprises requires some capital to purchase industrial sewing machines and good relationships with shoe manufacturers who entrust them with the very expensive leather. According to survey results, most working children under 15 (specifically, 95.8%) are employed by their parents on a part-time basis in these shops, which are often located close to the house and sometimes in it. Parents find it attractive to employ their children. First, as employers, they can safely put their underage children to work for them, pay them very little, and enhance greatly the flexibility of their operation by using them only when needed. As parents, they believe this is a desirable arrangement in that their children are learning skills of value and they are keeping them under their supervision and away from the temptations of the street (Licursi, 1995).

Thus, any analysis of the causes of child labor in Franca's shoe industry must begin with employers, the production system they have created and the nature of their labor demands. First, the production system, based on subcontracting to family-run shops often operating within the house, clearly facilitates child labor in forms that are virtually impossible to eradicate. Second, given the context of extensive unemployment, employers have the power to be highly selective in their choice of workers. As a result, one would expect that children working in the shoe industry would come from families with more resources and better connections to employers than many of those not working.

Local cultural traditions have, like poverty, often been advanced to explain child labor (UNICEF, 1998). In Brazil, many attribute the prevalence of child labor to the high value traditionally placed on work. Abreu and Martinez (1996) argue that this ideology developed in the late 1800s as movements to free slaves led to concerns with work motivation and discipline of soon-to-be-free labor. As elaborated, this pro-work ideology appears to embrace several notions. However humble, work dignifies those who perform it. Additionally, work serves as a school for life and is thus often more valuable than formal education. And third, work keeps individuals away from the temptations and dangers they might encounter and succumb to if they were idle. The corollary is, of course, that children should start doing some work in some form at an early age.

Many adults in Franca articulate this ideology in efforts to explain and justify work by children. Moreover, many of the children echo the adults’ sentiments. In the survey, over 80% of children and adolescents indicated that learning was an important work motive for them and 95% felt that getting a good job required work experience. Over 90% also indicated that they worked because work was far preferable to "the street." By this, they imply that they have no other legitimate activities to occupy their time and that continued idleness might lead them to leave the safety of their home for the excitement and dangers of public streets. Finally, in the interviews, many argued the importance of starting to work early in order to make something of themselves.

Nonetheless, what appears far more consequential than the traditional culture for decisions to seek work is the global youth culture of consumerism. When asked about their reasons for working in the survey, over 80% of the respondents reported that earning money for themselves was important. The salience of money was even more evident in many of the interviews conducted with young workers in commerce as well as the shoe industry. All were asked the open-ended question, "Why are you working?" The most frequent response by far, volunteered by about 55% of the shoe workers, was essentially that they wanted to earn money so they wouldn't have to depend on their parents to buy them what they wanted. Most use their money to buy branded clothes and shoes, CDs and CD players, and tickets to movies, concerts, fairs, rodeos, and the like. In addition, girls report lots of cosmetics purchases. Many report that they have often argued with their parents about their needs for these items and to avoid these conflicts, they decided to look for work. Two quotes give the flavor of the respondents' remarks:

Underlying these conflicts has been the diffusion of global youth cultures to developing areas. Since at least the 1970s if not before, these cultures have been carried through the mass media (most notably television) to virtually every nook and cranny in Brazil. As shown earlier, virtually all shoe workers' families have television. Data from a survey question about their viewing habits indicated that about 60% of both the children and adolescents watch 6 or more hours per week. More recently, the opening of Brazil’s economy has made many foreign consumer items newly available via imports or local production by foreign firms. Increasing competition within the domestic market has unleashed the energies of advertising with now familiar efforts to segment markets and appeal directly to teenagers and pre-teens. The drastic reduction in inflation in the mid-1990s has radically increased the availability of consumer credit. Symptomatic of the consumer revolution taking place in Brazil has been the rapid increase in the construction of shopping centers throughout the country. Franca’s first shopping center, located on the periphery of this city, opened in 1995 and features the numbingly-familiar gamut of restaurants, movie theaters, and stores specializing in shoes, clothes, and consumer appliances. Nearly 30% of the young shoe workers and 20% of the adolescent workers reported that they frequently went to the shopping center with less than 5% in each case saying that they had never been there.

Thus, the most identifiable causes of child labor in the shoe industry can be located not in characteristics of Brazil’s traditions or its "backwardness," but rather in the ways in which this nation is becoming integrated economically and culturally in the global economic, political, and cultural systems.

The implications of child labor in Franca for the children, their families, adult workers, and the society at large have been hotly debated since the mid-nineties when child labor became a public issue. At that time, the results of research by the shoemakers union revealing extensive child labor in Franca's shoe industry were widely disseminated both within Brazil and abroad (Licursi, 1995). International agencies such as UNICEF and the ILO as well as national ministries of labor from the US and Western European nations sent representatives to investigate.

Despite sharp income and wealth differentials in Franca, many of the more affluent classes owe their success to hard work in shoe factories, feel a sense of kinship with the less affluent workers, and participate in a variety of charitable activities designed to ensure the welfare of all. As a result, many deeply resented the stigma attached to their names and that of their city by the widely-circulated charges of exploitative child labor. That the local shoemakers union had led the attack served to heighten their anger and made even easier their denials. Throughout the 1980s, many felt themselves to have been bloodied by strikes and other encounters with local leaders of this militant group that formed part of the much larger, left-wing union federation (Casemiro, 1998).

For many adults, work by children in the shoe industry is seen as normal, benign, or healthy for their development. Typically, they make three points. First, they argue that children are rarely forced to work by their parents. Most of them allegedly want to work and enjoy it. In the survey, children substantiate this position to some degree with 64% of the 12-14 year olds and 59% of the adolescents denying that they worked because their parents asked them to do so. Second, many adults point to the fact that most young children with jobs work for their parents or relatives (over 95% of 12-14 year old shoe workers in 1998 sample) and not for impersonal firms. The implication they draw is, of course, that parents are responsible and do not exploit their children.

Third, they note that the vast majority of working children also attend school on a regular basis. This appears to be generally the case in Franca and, indeed, throughout Brazil. Data from the surveys and interviews suggest that young shoemakers not only attend school, but also regard formal education as critical to their ability to "ser alguém na vida" or make something of themselves. About 95% of the 12-14 year old shoe workers believe education is very important for their career success and most hope to attend college. Moreover, comparisons, via analysis of variance techniques (see Table 3), of the young shoe workers with three other groups of students (those working in commerce and services, those not working but looking for work, and those not working by choice) indicate that they are as involved in the school and as successful academically as the others.

The concerns that young people express regarding their education appear to reflect those of many parents as well and have been nurtured by continuing efforts by employers within and outside the shoe industry to raise the educational credentials required to be hired and promoted. Shoe manufacturers have taken the position that they can only hope to compete in world markets by constant increases in the quality of their shoes that, they believe, necessitates a more skilled work force.

The conventional wisdom notwithstanding, other information leads to a more skeptical view of work by children and adolescents in Franca. First, while it appears true that many children "choose" to work in the shoe industry, their choices are severely circumscribed. Children in public schools attend class in shifts of 4-5 hours per day. Because public agencies provide few after-school activities for school-age children, they have a great deal of time on their hands. As shown earlier, when asked why they work, almost all indicate that a very important reason is that they had nothing better to do ("work is better than the street"). Under these circumstances, the household-based subcontracting system makes it easy and natural for young children to join older siblings and parents in another room and "help out" with the work. As indicated earlier, at least 1200 of these bancas function in support of the manufacturers. Older children who lack access to jobs through their parents frequently look for part-time jobs to earn some money of their own to buy things. Because shoes dominate Franca and account, directly or indirectly, for nearly 50% of all jobs (Franca City Hall, 1999) they naturally drift into shoe jobs because these are more readily available and closer to their homes. Thus, short school shifts, few after-school activities, shoe industry dominance, and household production channel and guide young people to jobs in small bancas and shoe factories.

Second, the presumption that work for parents is somehow safer and more beneficial to young children than work for unrelated adults may be challenged. Most importantly, shoe production processes use glue to bind leather pieces together to facilitate stitching. Much of this glue is toxic, and young and growing children are more susceptible to neurological damage as a result of prolonged exposure. Family enterprises are not less prone to expose young workers to toxic glues than (nonfamily) firms. Over a third of the 12-14 year old children employed by their parents report that they work with glue as do similar portions of adolescents working for their parents and for firms. In addition, young children working in their parents' bancas or factories typically are paid far less (R$63.5 or about US$36 per month: R$.58 per hour) than adolescents working for families (R$160.4 per month; R$1.05 per hour ) or firms (R$ 173.1 per month; R$ 1.04 per hour). Given the very simple jobs that most young workers perform, it is difficult to imagine that productivity differences can justify these differentials. More likely, parents can exploit their young children in this fashion because they are underage and cannot get work outside the family. Adolescents, by contrast, have options outside the family business and often do leave to earn more elsewhere.

Third, families appear to discriminate against female children in Brazil’s shoe industry. Most of the young children working in family shoe enterprises are boys (69.5%) and, for this, they earn some money, although not a great deal. Young girls work as well, but doing unpaid chores for the family, oftentimes to release older females to work outside the home. Because their duties often include watching younger children, their hours of domestic work often exceed those of their brothers in the bancas and, according so some, are likely to interfere greatly in their school activities. When girls are allowed to work for their families, they are far more likely to labor as "coladeiras" or glue workers than their brothers. Of the 15 girls in the 1998 sample, 12-17 years of age, working for parents, 9 or 60% were coladeiras spreading glue on leather pieces. By contrast, among the boys working with their parents only 6 of 31 or 19% performed a similar function. Finally, as shown in Table 4, girls consistently earn less per month and per hour within family enterprises than their brothers. Indeed, the gender pay gap appears larger in family businesses than in firms.

Finally, that young shoe workers continue to attend school and perform as well as non-workers may say more about deficiencies in the educational system than about the benign nature of shoe work. Brazilian authorities have been widely criticized for the failure to invest sufficiently in public elementary and secondary schools. This failure is patently evident in more highly developed regions such as São Paulo and Franca. As indicated, schools do not have space to accommodate the rapidly expanding population so they function on shifts of 4-5 hours per day. Classes are usually very large with 40-45 typically assigned to a single, poorly-paid teacher who may teach in 2-3 schools on any day. Many teachers, especially the younger ones, appear to lack training and virtually all struggle to manage the large numbers without benefit of adequate support personnel, classroom supplies, library resources, computer and science labs, and recreational facilities (Madalena, 2000).

The fifty-minute classes often begin with teachers writing out lessons on the blackboard. Subsequently, they lecture the class and ask that the lesson be copied in their notebooks. During many classes, small groups of boys disrupt the lectures while others converse with one another, listen to their Walkmen, or gaze out the window. Many teachers seem to believe that the only reason at least some students remain in school is because employers demand diplomas. Girls that were interviewed often complain about their inability to learn anything because of the lack of interest and disruptive behavior by the boys. Many teachers lack training in classroom management as well as the ability to sanction through grades. Typically, little homework is assigned, students reportedly find it very easy to cheat on tests, and even if they perform poorly, policy changes implemented two years ago now prevent elementary schools from holding them back. Virtually automatic promotion has been adopted because high failure rates in the past created undesirable mixes of ages within classes and pressures to have night classes for elementary classes so some of the older elementary students could work.

The deficiencies, practices, and policies in Franca's schools suggest, first, that children with part-time shoe jobs remain in school and do as well as non-workers not because of the benign nature of parent employers, but rather due to the lack of performance demands made of them by the schools. Also, the poor quality of schools may well reinforce the traditional bias in Brazil in favor of work over schools and lead some students to seek jobs. At the minimum, some students choose to work more hours because school is so easy and irrelevant.

Much of the discussions of child labor focuses on the effects of early work on the children themselves. It is important to recognize that the implications of child labor extend to other groups. The greatest secondary impacts sometimes fall on families of child workers. If, as we suggest above, adolescents in Franca do not work out of economic necessity, then the loss of their contribution to family income does not seem to be a great economic hardship. Given that a substantial percentage of their income is probably being used for personal consumption, that contribution to family income may be minimal anyway. On the other hand, adult workers are likely to be hurt by the labor force participation of adolescents especially when, as in the present case, unemployment is very high. Because they provide a cheaper and more controllable alternative to adult workers the employment of child workers is likely to raise even higher the unemployment of adult workers. More generally, child labor constitutes a part of the informal economy that undermines governmental oversight of economic activities and therefore the rights of workers under the extensive labor codes. Legal constraints on employers within Brazil are especially important to labor given the abundant labor supply, underdevelopment of collective bargaining, and lack of within-company union representation that, in other countries, performs a monitoring function.

Responses to the issue of child labor have taken several forms in Franca, involved different groups and constituencies, sought different goals, and justified their actions on different grounds. The most visible response has come from the shoe industry. Under intense US pressure, the shoe employers association, dominated by large, export manufacturers, has developed a series of programs and procedures at least ostensibly designed to eradicate child labor from the industry. To administer these programs, the association founded the Pro-Child Institute (PCI) and urged its 200 members to join and pledge to eliminate child labor both in their own factories and among their subcontractors. Those joining the PCI (about 50 large manufacturers accounting for about 70% of shoe production in Franca) agree to submit to unscheduled audits by members of a local business college and to make their subcontractors available for such audits as well. Participating manufacturers are inspected bi-annually as are 10-15% of their subcontractors, chosen randomly. A first offense results in a warning with a second resulting in the firm or banca being expelled from the PCI program.

Participating firms fund this program by paying a fee set on a sliding scale from about U$25 to US$200 per month. In return for their participation in PCI manufacturers enjoy the right to place the PCI label on their shoeboxes. Of far more importance, however, is the legal defense they can make when and if faced with Ministry of Labor charges of child labor use among their subcontractors where the vast majority of child laborers are found. In effect, firms can claim that they have no role in the matter because they contracted only with bancas promising not to use child labor, required that their subcontractors submit to surprise audits, and helped finance the audits (Casemiro, 1998).

In addition to these activities, the PCI also works to develop after-school activities for children through partnerships with a variety of local firms in Franca. These include short courses in auto-mechanics, English, computers, music, arts, sports, and typing. At present, officials are developing an apprenticeship program that will enable 14-15 year old sons and daughters of already-employed shoe workers to learn a variety of skills through supervised instruction in shoe factories while receiving classroom training through SENAI, the national employer associations training arm. However useful these activities, it is clear that the primary mission of the PCI has been to defend the interests of the shoe exporters through serious efforts to eradicate child labor in their members' production networks. How well they have done this is difficult to say. By all accounts child labor has been virtually eliminated in these manufacturing plants of its members.

The impact of PCI on child labor in bancas is harder to ascertain. This is because many of these are home-based operations and detecting child labor violations in the home is very difficult. Because the penalty for using child labor is severe there is a strong motivation to hide this activity and doing so is quite easy. A child, for example, might be helping his parents with sewing operations. When the inspector appears at the door the child could very easily put down the shoe he/she is working on and sit down to watch television, read or engage in any of a number of activities. Thus the likelihood that an inspector would actually catch a child working in a PCI member banca is very slim.

There is another concern about the overall impact of PCI and, indeed, all such programs that focus solely on an industry rather than labor market. That is, to the extent that the problem of child labor is ameliorated in an affected industry, child workers may simply move on to work in other industries. This is especially likely to be the case if the underlying causes of child labor are not addressed. To make matters worse, this new work may be even more dangerous, particularly if opportunities for work in legitimate industries are cut off. Children then resort to illegal activities such as prostitution, the drug trade, or theft.

Two additional entities in Franca have developed programs that have the potential to complement PCI efforts to eradicate child labor in the shoe industry by absorbing some of the children and adolescents allegedly displaced by the PCI efforts. As a result of the union research revealing extensive child labor in Franca, UNICEF, with assistance from German donors, helped to develop a program that pays about 180 poor families with young children stipends of about U$40 per month to keep their 600 children out of the labor force and within school. Subsequently, this program came to be funded and administered through the local government when the Workers Party seized control in the late nineties. Recently, the federal government has developed two similar income-support programs that promise to cover larger numbers of the population. The first, called the Program to Eradicate Child Labor, is run out of the Ministry of Social Security and provides about US$15 per month per child in the 7-14 year range and reaches 87 families. The second and larger, called School Scholarship, is operated by the Ministry of Education and will cover 2000 low-income families, but provides only about US$6.00 per month per child in the 7-14 year range. All these programs are aimed at families with per capita incomes of less than half a minimum salary or about US$35 per month (Licursi, 2001).

The provision of financial incentives to withdraw child workers from the labor force could obviously complement the PCI programs by enabling families with young shoe workers to maintain their incomes while pulling their children out of bancas and shoe factories. The extent to which this occurs depends to a considerable degree on the ability of families with young shoe workers to demonstrate that their incomes, without any funds supplied by working children, are sufficiently low to qualify for these programs. In many cases, these families would not be able to do so since typically both the husband and wife are employed, often in the shoe industry. Thus unless some special efforts are made to enable shoe families to receive this income supplement, one would suspect that the program will have relatively little effect in facilitating the flow of child workers out of the shoe industry.

Another very significant program, founded and run by the Rotary Club of Franca, places about 390 young people, 14-16 years of age, in paid apprenticeships with area businesses. Although qualifying under Ministry of Labor guidelines as an apprenticeship program, the objective is not to develop marketable skills among the participants, but rather to contribute to their general intellectual and moral development. Competition to enter the program is keen with about 1600-1800 applying for one of the 200 positions opening up each year. Critical to the selection process are the results of written examinations on several academic subjects. Once admitted, the apprentices are assigned to businesses where they typically perform clerical duties. When insufficient businesses are available, some entering boys are sent initially to monitor parking in the downtown area. In addition to working 35-40 hours per week, they are required to continue their studies in evening programs, maintain good grades and comportment, and to participate in short vocational courses (usually 2 hours on Saturdays) offered at the headquarters of the organization. The apprentices receive about the minimum wage (currently around US$70 per month) plus many benefits and many apparently remain with their sponsoring firms on a permanent basis (Leal, 2001).

As with the income support program, this apprenticeship program could support the PCI efforts by facilitating the transfer of underage workers out of the shoe industry and into these legal and regulated forms of office employment. However, the intense competition for positions and the use of academic exam scores for selection suggest that middle and upper middle class children find it far easier to secure apprenticeships than working class children employed in the shoe industry. However, it is, of course, easy to conceive of changes in the admittance criteria that would make it easier for young shoe workers to enter.

Modifications in the income support and apprenticeship programs to facilitate the work of the PCI appear unlikely to be made because of the history of antagonism between the three program sponsors. Shoe industry officials in charge of the PCI remember keenly and deeply resent the leading roles played by officials currently associated with the income support program in the revelation of child labor in Franca. Reflecting past conflicts, the latter remain suspicious of PCI motives, routinely belittle the impact of their programs and note the export successes of the large firms. At the same time, both parties frequently criticize the work of the Rotary Club alleging that the so-called apprenticeship program provides little training and is really a form of disguised child labor. Rotary personnel, for their parts, are exceedingly sensitive to these charges and vehemently deny them.

The Franca case holds many implications for efforts to understand and respond to child labor in global industries. First, the most salient causes of child labor in Brazil's export-oriented shoe industry appear to have less to do with underdevelopment or the persistence of traditions and much more to do with the way in which this industry and its workers are integrated in the emerging global order. Although many factors play a role, the diffusion of consumerist cultures to Brazil stimulates needs in children while a production system, organized in response to competitive pressures and based on subcontracting to household units, enables them to evade legal prohibitions, obtain work, and earn money to meet those needs. Second, the study also suggests the need to more closely examine child work for parents in family businesses. Assumptions that such work is invariably benign are not supported in the present case where families routinely subject their child workers to toxic glues, pay them far less than warranted by their productivity, and discriminate against female children by limiting their access to the better jobs within family businesses.

Third, this study suggests the need to rethink the relationship between education and child work. Certainly in global industries facing quality as well as cost pressures, the Franca phenomenon of extensive part-time work by young students is apt to become more common as employers increasingly favor those with higher levels of education. However, to the extent that public policies favor private over public investments, the practice of combining work with study may rest on a very degraded and undemanding public educational system that masks work-study conflicts. Such systems are, in turn, apt to discourage students and parents and, consequently, push children into work activities prematurely. Thus, while child labor may, under some circumstances, affect school performance, it is equally the case that characteristics of schools affect the incidence, intensity, and nature of child labor.

Fourth, efforts to eradicate child labor from single industries have the potential to simply displace young workers to other industries and sources of income. In some cases, the net result may be a decline in the welfare of the children involved as they are forced into the less regulated informal economy or the underworld of drugs, prostitution, and criminal activity. As a result, more integrated, community-level approaches that cover the relevant community and labor market are needed. Such approaches necessarily seek to prevent the displacement of child laborers from one industry to the other by addressing the host of factors that facilitate or encourage work by young people: household-based production systems, extensive marketing activities designed to create young consumers, child and family income needs, deficient and undemanding educational systems, inadequate after-school activities, and the lack of day-care facilities. However, the process of raising to public awareness the issue of child labor may, as in the case of Franca, create deep antagonisms between those providing work for children and/or benefiting from it and those objecting to such practices. Such antagonisms may prevent the development of the host of cooperative programs at the community level needed to deal effectively the issue of work by the young. Thus, perhaps the most important role for international organizations is to work with and support local program sponsors to build upon and integrated their activities with those of other institutions seeking to promote similar ends.

Franca City Hall. 1999. City of Franca. Franca, Brazil: Gabinete do Prefeito

Glaser, B. G. and A. L Strauss. 1967. Discovery of Grounded Theory. Chicago: Aldine

Harkin, T. 1999. The United States Should Ban Imports of Products Made by Children. In

Hertel, S., with Powell, J., & Chowdhury, S. 1997. "Child Labor and International Sourcing."

Corporate Action Update. Council on Economic Priorities, March: 1-4.

Kinder, Lydenberg, and Domini. 2000. http://www.kld.com/ (accessed January 27).

Horn, P. 1994. Children's Work and Welfare, 1780-1890. Cambridge: Cambridge

University Press

(accessed January 11, 2000).

International Labour Office. 1998. Child Labour in the World. Geneva.

http://www.dol.gov/dol/ilab/public/media/reports/iclp/sweat5/chap2ft.htm

(accessed January 24, 2000).

Nader, R. 1996. "No child should work." USA Today. June 13: A7.

Nichols, M. 1993. Third-World Families at Work: Child labor or Child Care? Harvard

Business Review, January-February: 12-23.

Note: School Involvement is the mean of responses to three items asking respondents to indicate how interested they were in school, how much they liked it, and how satisfied they were with their progress. Each item had four response categories. The alpha coefficient is .70.

Note: Figures in Brazilian reais: exchange rate in 1998 of about R$1.75 to US$1