A. Introduction

Even though schooling is a potentially profitable investment in most countries around the world, many children still work. This chapter explores the factors that lead children into the world of work and identifies the barriers that prevent them from moving from work to school. As the chapter illustrates, there is no single cause or set of causes that can explain child labor in all contexts or in all countries.

This chapter does not aim to identify all the possible causes of child labor, but rather to discuss some of the most commonly identified causes. In doing so, it groups factors into three categories: a poverty of resources, a poverty of opportunities, and the availability of work.

B. Poverty of Resources

Poverty seems to be most commonly identified as the catch-all cause of child labor.¹ A “poverty of resources” is defined here to mean those instances when child labor is thought to result from low adult wages, low family income, or lack of financial assets.

The logic linking child labor to poverty is clear. Some children work because they or their families could not survive without the income, goods, or services generated by the work of children. Lower income families spend a higher percentage of their total income on essentials like food and shelter, and in many cases, may depend on the earnings of children to provide for these basic needs. In these families, not having their children work may be a luxury which is sought only after survival of the child and the family is ensured.²

1. Evidence

a. Country Level

Evidence supporting the proposition that child labor is linked to poverty is available at the country and at the household levels.³ In general, the poorer a country, the higher the incidence of child labor. The labor force participation rate of 10-14 year old children⁴ is between 30 and 60 percent in countries with gross domestic product (GDP) per capita of $500 or less (in 1987 dollars) and 10 to 30 percent for countries with GDP per capita in the range of $500 to $1000. Above per capita GDP levels of $1000, the incidence of child labor declines as GDP per capita increases, but the rate of decrease is less marked.⁵

The fact that the incidence of child labor declines less rapidly in countries with per capita GDP above $1000 is perhaps explained by who is poor. In these countries, a positive relationship may be seen between the incidence of child labor and income inequality: the more unequal the distribution of income in a country, the higher the incidence of child labor.⁶ More unequal distributions of income generally imply a greater disparity in the living standards of the “haves” and “have nots.” This evidence suggests that even in countries that are not extremely poor by measures of average household or individual income, there may be households subsisting far below the average. In these households, child labor may still be a reality.

b. Household level

Within a household, the likelihood that a child works depends on other sources of income available to the family and the number of people for whom those income sources must provide. Higher parental income reduces household pressure to send children to work and makes schooling alternatives more affordable. Parents who work full time are better able to provide for basic needs than parents who work only part time or on an irregular basis, and are less likely to be reliant on the earnings of their children.⁷ Conversely, when household income declines due to an adult’s falling wages or reduced hours, children in that household are at greater risk of early employment.⁸ In such instances, children are often forced to forgo schooling in order to supplement adult income in a household. A study in Brazil found that 48 percent of children who worked in order to supplement family income had dropped out of school early, as compared to a ten percent drop-out rate for all children in Brazil.⁹

i. Mothers as a source of income

Child labor can be particularly important in households where a parent is absent or deceased. The loss of a mother has been found to have a greater impact on children leaving school prematurely than the loss of a father. Moreover, the loss of a mother tends to have a particularly negative effect on girls who are frequently called upon to assume the domestic responsibilities previously carried out by their mother. In addition, mothers can provide a source of income that reduces the pressure for children to work. A rise in women’s earnings has been found to have a significantly positive impact on children’s education. One study in Egypt found that a ten percent increase in women’s market wages led to a 27 percent decline in employment among children between the ages of six and 11.¹⁰

ii. Family size

In households with large numbers of children, if income is insufficient to meet basic needs there will be pressure to send at least some children to work in order to supplement overall household income.¹¹ In some cases, however, being from a large family may provide children with the opportunity to go to school. A study in Botswana found that, on average, children from larger families actually are more likely to be enrolled in school and to complete higher levels of schooling. The study attributed these findings in part to the diminishing returns from each additional child’s entrance into the workforce.¹² There may also be a kind of “specialization” within the household, whereby larger numbers of children allow some to attend school because of the contribution to household income made by those working.¹³

2. Household decisions about child labor

In very poor households, it can be argued that there is little or no choice about whether or not children work. Children must work to survive.¹⁴ But the argument that these households face no other choice except to send their children to work probably holds true only for an extreme level of poverty.¹⁵ The more nuanced notion that poor families weigh what they give up by foregoing child labor against what they get in return seems more generally applicable.

The income that children provide to a poor household can be important. For example, children who work in Colombia contribute, on average, about 19 percent to the total income of their households in urban areas and 35 percent in rural areas.¹⁶ In urban Bolivia, working children aged seven to 12 years old contribute, on average, nearly 20 percent of the family income.¹⁷ The contribution generated by children is not restricted to cash income. In-kind income—direct goods or services provided by the child for consumption of the family—can also be very important. The Ugandan Government reports that within its country, “Family labor shortage in subsistence agriculture and lack of access to amenities such as water and firewood and the absence of energy saving devices, are some of the major causes of child labour in Uganda.”¹⁸ In other words, children fetch water and wood, and do chores that no one else is available to do, or for which machines to speed the work along are not available.

In addition to the income forgone if children attend school rather than work (i.e. the “opportunity costs of schooling”) households consider whether they can afford the out-of-pocket costs of education. The variety of these costs can be illustrated with an example from the Philippines, where despite tuition free education, families must pay for school supplies, uniforms, materials for projects, additional books, and contributions for special projects and activities, and in many cases, transportation.¹⁹

Meeting the costs of schooling is a particular challenge for poor households. High fees in combination with falling income have been related to low school enrollment and high levels of child labor.²⁰ A study in Ghana found the high cost of school- ing, coupled with low school quality, greatly increased pressure to send children to work instead of school.²¹ For many families in Zimbabwe, the direct and opportunity costs of schooling are prohibitive.²² Among children who live in urban Bolivia, those living in areas with higher costs of schooling are significantly more likely to work than those where schooling is cheaper.²³

Sometimes children work to finance the costs of schooling. In Zimbabwe, some children work in exchange for the opportunity to attend school;²⁴ they are required to complete a minimum amount of work or risk being withdrawn from school.²⁵ A study in Peru found no apparent trade-off between child labor and school enrollment. The authors concluded that working may provide the resources that make it possible for children to go to school.²⁶ For child labor and schooling to be complementary, however, children must have enough time and energy to attend and succeed in school. This suggests that there must be a limit to the time children spend working.²⁷ If working leaves children with insufficient time or energy to devote to studies, child labor has a negative effect on schooling.

3. Child Labor as Family Insurance

Another way that poverty affects how households make decisions regarding the allocation of children’s time is by influencing a family’s strategy for dealing with unanticipated interruptions in the earnings of its members.²⁸ Loss of income because of a poor harvest or the loss of work of a family member because of dismissal, injury, or sickness is a significant threat to families whose ability to provide basic necessities is marginal. This vulnerability to risk makes the short-term returns of child labor more attractive to lower income households, as long as the interruption of one family

member’s income can be somewhat offset by the others.²⁹ Working children, in effect, provide a means for diversifying risk. The more members in a household that work or can work, and the more diverse the sources of their income, the less vulnerable the family is to the loss of income from any one member.³⁰ Poorer families also generally have fewer income generating assets, such as a cattle, that would allow them to reduce their risks.³¹

Child labor is apparently used as a risk management strategy by rural households in India. Broad fluctuations in household income, and hence more chance of big disruptions of income, are associated with lower incidence of school attendance than narrow fluctuations.³² The notion of child labor as a form of insurance is also consistent with evidence that very poor households in Cote d’Ivoire, during a recent recession, increased the participation of secondary earners, primarily children.³³ For similar reasons, concerns about increased child labor were raised by the financial crisis that affected Southeast Asian countries in 1997 and 1998.³⁴

4. A Cycle of Poverty and Child Labor

From Chapter II, it is clear that children who do not attend school earn less as adults than children who do. Child labor that completely or partly displaces schooling thereby imposes this cost on children who work. But the effect of not educating one generation of children also tends to have a costly effect on the incidence of poverty and child labor in the next, leading to a cycle of poverty that perpetuates itself.

Children can be affected in many ways by the level of schooling completed by their parents. More educated parents generally earn more and are better able to provide for their families, reducing the likelihood that their children will need to work. Better educated parents are also likely to place a higher value on education and to be more supportive of investment in their children’s education. Studies in Paraguay,³⁵ Cote d’Ivoire,³⁶ Colombia,³⁷ Bolivia,³⁸ and the Philippines³⁹ reveal that higher levels of educational attainment by either the father or mother, or both, lowered the probability of their children working, or raised the probability of their children going to school, or both.

Available evidence most often suggests that poorly educated parents are less likely to allocate their children’s time to school, and more likely to allocate it to work. In a sense, child labor can be traced back to the work versus education decision made on behalf of a child’s parents by the child’s grandparents, and perhaps even further. And since less educated persons tend to have lower incomes and wealth, a decision to send children to work made by one generation helps to make poverty more likely in future generations of the family. Through this mechanism, child labor can be seen as a contributing factor to a cycle of poverty.

C. Poverty of Opportunities

To this point, poverty has been portrayed in a very traditional way—as a lack of financial wherewithal. In considering child labor, it is necessary to consider another sort of poverty: poverty that arises because the choices for children are so restricted that few options aside from working are available to them. This poverty of opportunities can be closely related to financial poverty. Financially poor families are likely to have or perceive few alternatives to work for their children. But it is important to make a distinction between poverty of resources and poverty of opportunities. While the generation of financial wealth may address the causes of child labor due solely to financial poverty, it may not be sufficient to address a lack of alternatives to working for many children. The elimination of financial poverty may also not be sufficient to ensure that opportunities are offered to all children.

This section discusses four issues related to a poverty of opportunities that affect many working children and their families: (1) a lack of appropriate schooling; (2) discrimination; (3) cultural attitudes; and (4) restricted access to credit.

1. Availability of appropriate schooling

Whether available schooling is appropriate depends on its accessibility, quality and relevance.

a. Access to school

Schooling is not an option for a child if it is not accessible.⁴⁰ If available schools are too far from where a child lives, the child’s family is not likely to consider schooling as a feasible use of the child’s time. For every additional kilometer that a Nepalese child must walk to school, it has been estimated that the likelihood of school attendance drops by 2.5 percent.⁴¹ In rural areas of Cote d’Ivoire, children are more likely to attend school if it is located in their village rather than far away.⁴² The same is true for rural Ghana, with the probability of school attendance declining with distance from a school.⁴³ A survey in Zimbabwe’s mining regions found that only those children who live near schools attend.⁴⁴ Low population density in rural areas and long distances to schools often mean that there are few alternatives to child labor in these areas.⁴⁵

A survey of 1,221 rural Indian parents found that about 50 percent of rural Indian out of school working children worked fewer than 3 hours on the day preceding the survey, while only around 18 percent worked 8 hours or more. The analysts conducting this study concluded that the light work load carried by these children should make school attendance possible if appropriate schooling were available to them.⁴⁶ Children in rural areas of Cote d’Ivoire may face a similar phenomenon. In comparisons among rural regions of Cote d’Ivoire, the Savannah region was found to have the highest incidence of working children and this higher incidence was linked to the fact that the “educational infrastructure in Savannah lags far behind the rest of the country, as it has for generations.”⁴⁷ A similar conclusion holds in a study of the Philip- pines, where the improvement of the availability and quality of schooling is identified as a particular need in rural and remote areas.⁴⁸

b. School quality

Where schools are available but education is of poor quality, children also face a lack of real opportunity because their education is unlikely to give them the skills and competencies needed to command higher wages in the labor market. With the pay-off to education so restricted, working may be seen as a better use of children’s time.

Low quality schooling can manifest itself in many ways. In many countries, classrooms tend to be roughly constructed with rooms that are poorly lit and inadequately equipped. Overcrowding, especially in urban areas, is also common. In Bangladesh for example, teachers are reported to have as many as 67 pupils in a classroom, while in Equatorial Guinea there may be as many as 90.⁴⁹ In India, 28 percent of schools have only one teacher who teaches an average of three to four classes.⁵⁰

Another feature of poor quality education is inadequately trained teachers. In Uganda, the Ministry of Planning and Economic Development found that 30 percent of the country’s teachers were untrained.⁵¹ Evidence from Egypt suggests that an overall increase in school quality has a major impact in terms of retaining children in school,⁵² and that low quality schooling contributes to a perception that schooling is ineffective.⁵³

c. Relevance of schooling

Schooling may not be seen as a real opportunity for children if it is not perceived as being relevant. Families will want to send their children to school if they see the potential for education to result in higher earnings later in life. But if better paying jobs that require the skills derived from schooling are few, parents are unlikely to see the value of investing in formal education. For example, in Kenya the formal economy is estimated to be unable to absorb even one quarter of the country’s secondary school graduates.⁵⁴ Graduates of primary school in Uganda are believed not to be equipped with practical skills needed to raise their earnings potential.⁵⁵ If the payoff from education in terms of better, higher paying jobs does not exist, the incentive for children to go to school is weak.

2. Discrimination

For girls and children from certain ethnic and social classes, discrimination may exacerbate the costs of child labor by further restricting their access to educational opportunities.

a. Gender

Gender plays a major role in determining opportunities available to children within a household. The available statistics on child labor around the world show that in most countries young boys are more likely to be classified as working than girls.⁵⁶ In fact, in all but two of the countries for which data are presented in Appendix B, boys were found to be more likely than girls to be classified as economically active.

But one of the major problems of child labor statistics is that not all forms of work are counted as “economic activity.” In most cases, domestic work that could be considered child labor is usually excluded. Since girls are much more likely to perform that type of work than boys, they may be undercounted in estimates of child labor. In two of the ten countries presented in Appendix B, Nepal (Table B-5) and Turkey (Table B-9), data were available showing the extent to which boys and girls performed domes- tic chores separately from their economic activity rates. In both cases, boys were more likely to be economically active, while girls were more likely to perform domestic housekeeping/homemaking services. Taken together, with “working” defined to include economic activity and housekeeping, girls were more likely than boys to be working. Likewise, in Tanzania where domestic activities were included in economic activity rates, girls were more likely to be economically active than boys (Table B-10).

Girls are often viewed as essential to the running of households in developing countries. When mothers work outside the home, girls frequently take responsibility for domestic work.⁵⁷ Girls are also often called upon to assume household responsibilities in the case of an ill or deceased mother. A study in Pakistan found that girls were less likely to attend school when there are younger children in the household, illustrating the important role played by girls in caring for younger siblings.⁵⁸

There is an interesting gender aspect in the relationship between parental education and child labor. The impact of parental education is particularly great for girls, since well educated parents are more likely than non-educated parents to enroll their daughters in school.⁵⁹

In some instances, mothers and daughters may function as complements in the labor market. In Cote d’Ivoire, an increase in the education of mothers was found to coincide with increased contemporaneous participation of their daughters in the work- place. Up to a certain level of schooling, more educated mothers who initiated enterprises tended to draw their daughters out of schooling to contribute to their new business.⁶⁰ Likewise, a study in India found that increased female wages led to a reduction in children’s schooling time.⁶¹ In general, it would be expected that at some level of education mothers would begin to command higher wages in the labor market and to be better positioned to invest in the education of all their children, including their daughters.⁶²

As a result of the roles played by girls within the household, gender can be a major factor in determining the educational opportunities open to children. In countries where patrilineal marriages are the norm, parents may be less inclined to invest in a daughter’s education since the resulting benefits are seen to flow from the family that invests in its daughters to the family whose son marries their daughter.⁶³ In some countries, education of girls is also seen to conflict with traditional notions of women’s role in the community.⁶⁴

b. Ethnicity or social class

Ethnicity or social class can be another dimension along which opportunities for some groups of children are restricted. In India, the caste system not only reinforces the roles of certain groups, but in effect, dictates the educational opportunities and likelihood of certain children working prematurely. A study of the granite and lime-stone industries of India’s Andhra Pradesh state found that religion and caste had an important influence over which children worked in these industries.⁶⁵ More generally, India has been portrayed as a society where child labor has flowed from the class biased nature of government education policy:

India’s policy makers have not regarded mass education as essential to India’s modernization. They have instead put resources into elite government schools, state-aided private schools, and higher education in an effort to create an educated class that is equal to educated classes in the West . . .⁶⁶

This belief system suggests that some people, even as children, are meant to work with their hands, while others are meant to work with their minds.⁶⁷

In Latin America, ethnicity also affects the incidence of child labor among certain groups of children. A child from an indigenous group is twice as likely to work as a child who is not of indigenous heritage.⁶⁸ In Bolivia, ethnicity also affects where children work. Working children from indigenous groups are more likely to be excluded from formal sector work, leaving for them mostly informal sector—and generally less desirable—work.⁶⁹

Language can also be a factor restricting opportunities for children from indigenous populations. In many of the English, French, and Portugese-speaking African countries, for example, lessons are still conducted in the former colonial language. This increases the chances that a child who speaks only the indigenous language will drop out of school, particularly when the child’s parents are illiterate.⁷⁰

3. Cultural attitudes that support child labor

Factors such as poor schools, discrimination, and financial costs can be insurmountable barriers for families who want to keep their children out of work and in school. These barriers can also help form or reinforce an additional barrier: a cultural attitude in the community that children should work and that they are better off doing so.

Child labor was once common and considered morally acceptable in many industrializing societies. Attitudes about work and childhood have undergone a great transformation since the late 18^th century, however, and in many countries, the notion that children should work has been rejected. Within the developing world where child labor remains prevalent, such attitudes about childhood and child labor often still persist.⁷¹

For example, a study in Colombia found that some adults continue to regard children as ‘mini-adults,’ therefore making them responsible for tasks that only adults in other societies would carry out.⁷² In India, a 1991 study documented a “near universal belief” among educated Indians that child labor is a ‘harsh reality’ of life among the poor, where children must work.⁷³

This attitude makes it more difficult for families to keep their children in school or to prevent them from working. They perceive work to be good for their children, a sound alternative education process. Even “non-poor” families send their children to work in the diamond industry in India because they believe these jobs are good training ground for their children in order to qualify for relatively well-paying jobs they may

obtain as adults.⁷⁴ One study in Ecuador also found that work was considered an educational opportunity for the young. Parents saw it as an opportunity for children to learn a skill or trade and to acquire a sense of responsibility. Parents defended child work as being instructive, teaching children to be responsible and to appreciate the value of things and the effort required to obtain them. Other parents believed that work enabled the young to learn an occupation or trade with which to support themselves as adults.⁷⁵ While in Brazil, one study concluded that some adults still consider child labor to be a part of the socialization process and a form of education.⁷⁶

These attitudes may be a reflection of other barriers a family faces. Concerns regarding the family’s economic options or the quality of the educational opportunities can lead families to the belief that their child is better off working than going to school. Regardless of its validity, however, this perception can become another barrier to be overcome.

4. Restricted Access to Credit

When seeking a loan, a borrower can go either to a formal lender, such as a bank, or to a variety of informal lenders, such as a landlord, a merchant, or an employer, whose primary activity is not lending money. In developing countries, especially in rural areas, poorer families whose children are at risk of child labor frequently have difficulty gaining access to formal lenders. Banks in developing countries often do not maintain networks in rural areas, leaving residents of these areas with access only to informal lenders.⁷⁷ Additionally, banks are less likely to lend small amounts of money to borrowers because small borrowers often have insufficient collateral; are unable to demonstrate an ability to repay loans; and may be seen as one-time borrow- ers who might be less concerned about failing to repay a loan since they have no intention of borrowing again in the future. Furthermore, legal means to punish small borrowers for nonpayment may be ineffective or nonexistent.⁷⁸

Because formal lenders are unavailable or unwilling to lend to poor, rural families, in many instances, the only option available to such families is to turn to informal lenders. To see what families are up against if they seek to borrow from such informal lenders, however, consider Table III-1. The table presents interest rates offered on loans by formal and informal lenders in a selected sample of countries. Informal interest rates range from 1.6 (Vietnam) to 33.3 (Nigeria) times greater than formal ones. As the table suggests, lack of access to formal lenders makes borrowing an expensive proposition and undoubtedly discourages it for many families who end up sending their children to work.

Restricted access to credit can be related to child labor in several ways: making it more difficult for families to find ways to afford schooling for their children; restricting families from pursuing possible income generating activities; and making extreme forms of indebtedness such as bonded labor more likely.

If families do not have the opportunity to borrow, they are forced to pay educational expenses for their children out of their current income or wealth. The more limited these sources, the less likely it is that their children will go to school and the more likely that they will work. If a family expects an investment in a child’s education to be profitable, they might want to borrow to cover schooling costs. Unfortunately, borrowing to pay educational expenses may simply not be an option available in many developing countries.⁷⁹ Furthermore, if families have to pay extremely high interest rates on loans taken out from the informal sector, it is unlikely that most will view education as a profitable enough investment to justify the expense.⁸⁰

If credit is not available, families cannot borrow to finance activities such as small businesses. Such enterprises are a potential source of future income for families that could allow them both to support themselves without sending their children to work and to pay the out-of-pocket costs of schooling. In fact, informal sector interest rates discourage many forms of investment. Lack of access to formal sources of financing has been identified as a restraint on small enterprise creation and growth in many countries.⁸¹ There is concern that women in particular lack access to financing.⁸² If families cannot borrow at rates that allow them to pursue profitable income-generating activities, one route out of poverty is closed off to them. Earlier in the chapter, low family income—particularly a mother’s income—was identified as a possible determinant of whether children work. It follows that any restriction of opportunities to generate income may also be related to child labor. Lack of access to credit, or access to credit only on terms that make it effectively unavailable, clearly represents one example of such a restriction.

Finally, if credit is available only on the most exorbitant terms, families that avail themselves of it may find themselves at risk of bonding their children into labor. The effect may actually be to ensure child labor rather than avoid it. Debt bondage occurs when labor is pledged in return for a loan.⁸³ Landless households in impoverished rural areas may become vulnerable if they are unable to meet their daily needs with their current resources. In some cases, parents may pledge their own labor or their children’s for an indefinite period of time. Lacking in education and negotiating power, they may agree to terms that include low wages, high interest rates, or both. The net effect is that it becomes impossible for them ever to pay-off the loan. Even if only the parent’s labor was originally pledged to work off the loan, the responsibility can eventually be inherited by their children and passed on to subsequent generations.⁸⁴ In situations such as these, lack of access to formal credit markets does not just keep families from making themselves and their children better off, it actually forces them into situations that make them worse off.

D. Availability of Work

The discussion so far has focused on why children are made available for work. But in order for child labor to exist, work must also be available for children. It is useful then to consider the factors that create work for children. Two questions are addressed: (1) do employers profit from employing children instead of adults; and (2) what is the role of technology in creating work for children?

1. Profits from employing children

One argument frequently made to explain the existence of child labor is that firms employ children because they are a “cheaper” source of labor than adults. While it is generally observed that children are paid less than adults, even when they perform the same jobs, it is not clear why.⁸⁵ There are two competing theories that explain why children’s wages are generally less: a) children’s productivity and quality of work are lower than that of an adult, or b) children are easier to exploit.

If children produce at a slower pace or at a lower quality than adults because of less skill, experience, education, stamina, or dexterity, the wages paid to children should be lower. If children earned as much as or more than adults, the firm would be better off employing only adults, who for the same cost, would produce more or at a higher level of quality. This suggests that child labor can only exist if the wage paid to children is lower than that of adults by at least enough to offset the productivity differences.⁸⁶ For example, if wages are paid on an hourly basis, and if under identical conditions, a child making a carpet can stitch half as many knots of identical quality as an adult in one hour, the employer would only hire the child at a wage equal to or less than half that received by the adult.

But lower wages are not the only factor explaining demand for child labor. Children may be compensated less for their work than adults in other ways. Children are often described as more compliant than adults in the workplace. They are less likely to complain about poor working conditions or to organize to improve them. Insomuch as this reduces an employer’s expenditure on workplace conditions, employment of children may be less costly. This argument suggests that even if children are equally productive, children will be paid less than adults. Another factor worth consideration is that absenteeism among child workers tends to be lower than among adults.⁸⁷ These factors increase incentives for firms or employers to hire children, but they also demonstrate the inherent danger to children of being exploited in the workplace.

Both of these explanations lead to the prediction that children will be paid less than adults. In the first case, firms are not necessarily exploiting children, but merely treating them as other low-skilled workers. In the second case, employers may prefer to hire children precisely because they can exploit them and earn more profit. While it is not possible with the current research to determine the extent to which either or both explanations are true, it is certain that simply analyzing the difference between the wages of children and adults does not show whether employers prefer to employ children.

2. Technology and Child Labor

a. Children as suited to certain forms of work

Usually, the superiority of children for certain jobs is identified as related to physical attributes such as size or agility. For example, boys are portrayed as superior to men for work in mines or as chimney sweeps because their size makes it easier for them to maneuver in small spaces. Similarly, children are portrayed as more desirable

than adults for the work of weaving high quality carpets because their “nimble fingers” make it possible for them to tie smaller, tighter knots.⁸⁸

Available evidence appears to go against the notion that child labor exists because children possess special attributes that makes them superior to adults for some types of work. For example, a study of the Indian carpet industry found that children did not weave a higher proportion than adults of carpets with the difficult designs that require tighter knots. Children, in fact, did not dominate any particular difficulty level of the industry’s carpet production.⁸⁹

Studies of other Indian industries reinforce the notion that children do not possess unique attributes that make them better suited for certain occupations or tasks than adults. In India’s glass industry, an argument is sometimes made that children are essential to production because their small size enables them to move about the glass factory faster and with greater ease than adult workers. A study of the glass industry in Uttar Pradesh, however, found that rather than being unique in the work they performed, children worked alongside adults. It demonstrated that children generally performed the lowest skill jobs, where worker substitution could most easily occur.⁹⁰

Children are also thought to be better suited for diamond polishing because the work requires acute eyesight; however, a study of India’s diamond industry also found children working alongside adult workers in all stages of diamond processing.⁹¹ In India’s gem stone industry, a study found that a majority of children performed un- skilled work where they could easily be replaced by adult workers, and in fact, many adults were engaged concurrently in these tasks with working children.⁹² In the granite and limestone industries of India’s Andhra Pradesh state, children perform mainly low skill, manual labor that could be done easily, and equally well, by adults.⁹³

If employment of children cannot be traced to children possessing unique attributes or skills, then working children should be considered to be part of the same pool of labor as adult workers. In the cases described above, children and adults function as replacements for one another. Children are not irreplaceable in these work environments. Rather, like adults, children are found working where their skill level allows them to contribute to the earnings of their employers.

b. Children as unskilled labor

Children may work because the prevailing organization of production requires a large pool of unskilled labor and the pool of available adult laborers is not large enough to meet this requirement. This is often the case in agricultural areas during peak labor seasons, such as planting or harvesting. In other cases, unskilled labor in the form of children may come as a sort of “package deal” when parents are employed in a job. Parents may bring children to work because they lack schooling or child care options and cannot afford to miss work to care for their children themselves. According to an ILO/IPEC survey of Southeast Asian manufacturing industries, employers explained the employment of child labor as driven, not by lower costs associated with children’s wages, but rather, by the relative abundance of child workers and the resulting ease involved in hiring them.⁹⁴

If child labor can be traced to a work organization that makes use of an abundance of unskilled labor, then technological change that replaces unskilled labor with machines or skilled workers should reduce child labor. During the industrial revolution in the United States and Great Britain, the introduction of machines for spinning and weaving brought about a decreased demand for child labor.⁹⁵ The green revolution in India brought about a variety of labor-saving changes in agriculture that led to a decrease in child labor and an increase in school attendance.⁹⁶ Similarly, a fall in child labor has been linked to the expanded use of tractors and modern irrigation techniques in Egyptian agriculture.⁹⁷ In the Philippines, the introduction of electricity in a community has been associated with a fall in child labor in market-based activities, and the availability of electricity in homes has similarly led to a reduction in time spent in home production.⁹⁸ And in Bogotá, Colombia, the use of children in the quarries was reduced after the introduction of wheelbarrows eliminated the need for carrying stones one at a time.⁹⁹

E. Conclusions

This chapter discussed different possible explanations for why children work. The purpose of this discussion was to lead up to an explicit identification of the barriers to the removal of children from work and their increased participation in school. It should be kept in mind that not all barriers operate in all contexts. The anecdotal nature of the evidence examined in this chapter suggests that child labor in one country may be traced to a totally different set of barriers than child labor in another country.

1. Barriers Related to a Poverty of Resources

Financial poverty, which was defined as a family’s inability to survive financially without child labor, is associated with a number of barriers to lowering or eliminating child labor and increasing enrollment of children in school:

• Inability of parents to support their families from their own earnings or wealth;

• Inequality in the distribution of income or resources;

• Lost income from children not working, or high out-of-pocket costs to their schooling, or both;

• The use of child labor as insurance against interruptions in the earnings of other members of the household; and

• A cycle of poverty within a family resulting from repeated generations of children working instead of going to school.

2. Barriers Related to a Poverty of Opportunities

Another set of barriers to the removal of children from work and enrolling them in school can be traced to a lack of alternatives to work for children or groups of children:

• Inaccessible schools;

• Low quality schools or education that is of little relevance;

• Cultural patterns that prevent or discourage the enrollment of girls;

• Attitudes suggesting that certain ethnic or social class groups are meant to work with their hands while others are more suited to working with their minds;

• Educational instruction carried out in unfamiliar languages that make it difficult for children to grasp the concepts conveyed; and

• Lack of available credit markets.

3. Barriers Related to the Availability of Work

Finally, there are barriers related to the fact that work is available for children to do, and that it would have to be done some other way if child labor were eliminated:

• Children might be “cheaper” to employ than adults because they are more pliable and less likely to resist poor working conditions, although it is still unclear whether or not this is true; and
• Production processes that do not rely on labor saving devices and/or an abundant pool of unskilled labor can create a demand for child labor.

From this chapter’s discussion, it is clear that there is no single set of causes of child labor that is operable in all contexts. Rather, the factors that create barriers to the removal of children from work and their enrollment in school can vary from country to country.

Having identified the barriers to moving children from work to school, it is natural to ask what can be done to lower them. The next chapter focuses on this question by describing policy strategies and initiatives that should reduce the incidence of child labor.

¹ See, for example, Child Labour: Targeting the Intolerable (Geneva: International Labour Office, 1996) 17 [hereinafter Targeting the Intolerable]; World Bank, World Development Report 1995: Workers in an Integrating World (New York: Oxford University Press, 1995) 72; Richard Freeman, “A Hard-Headed Look at Labor Standards” [hereinafter “Hard- Headed Look at Labor Standards”], in G. K. Schoepfle and K. A. Swinnerton, eds., International Labor Standards and Global Economic Integration: Proceedings of a Symposium (Washington, D.C.: U.S. Department of Labor, 1994) 31 [hereinafter International Labor Standards and Global Economic Integration Symposium].

² K. Basu and P. H. Van, “The Economics of Child Labor,” American Economic Review, 88 (3) (1998) 413 [hereinafter “Economics of Child Labor”].

³ There is a vast amount of empirical evidence relating poverty either to child labor or to low levels of educational attainment. This discussion gives only a cursory introduction to that evidence. References containing either more detailed literature reviews or more specific examples include: Shahnaz Hamid, “A Micro Analysis of Urban Child Labour: Some Determinants of Labour and Its Conditions,” Pakistan Development Review 33 (4) Part 2 (Winter 1994) 1249-69; George J. Mergos, “The Economic Contribution of Children in Peasant Agriculture and the Effect of Education: Evidence from the Philippines,” Pakistan Development Review 31 (2) (Summer 1992) 189-201; S. Canagarajah and H. Coulombe, “Child Labor and Schooling in Ghana,” Policy Research Working Paper No. 1844 (Washington, DC: World Bank,1997) [hereinafter “Child Labor and Schooling in Ghana”]; G. Psacharopoulos and H. Yang, “Educational Attainment among Venezuela Youth: An Analysis of its Determinants,” International Journal of Educational Development 11 (4) (1991) 289-294 [hereinafter “Educational Attainment in Venezuela”]; H. A. Patrinos, E. Velez, and G. Psacharopoulos, “Language, Education and Earnings in Asuncion, Paraguay,” Journal of Developing Areas 29 (October 1994) 57-68 [hereinafter “Education and Earnings in Paraguay”] ; Victor Lavy, “School supply constraints and children’s educational outcomes in rural Ghana,” Journal of Development Economics 51 (1996) 291-314 [hereinafter “Educational Outcomes in Rural Ghana”]. See also C. Grootaert and H. A. Patrinos (eds.), The Policy Analysis of Child Labor: A Comparative Study (New York: St. Martin’s Press, 1999) [hereinafter Policy Analysis of Child Labor].

⁴ Comparable and geographically comprehensive data covering labor force participation rates of children younger than this are not available.

⁵ P. Fallon and Z. Tzannatos, Child Labor: Issues and Directions for the World Bank (Washington, D.C.: The World Bank, 1998) 3. See also, Alan B. Krueger, “Observations on International Labor Standards and Trade,” National Bureau of Economic Research Working Paper #5632 (Cambridge, MA, 1996) 24.

⁶ C. A. Rogers and K. A. Swinnerton, “Inequality, Productivity and Child Labor,” Georgetown University Department of Economics Working Paper # 99-10 (Washington, D.C., 1999) 4. See also Priya Ranjan, “Credit Constraints and the Phenomenon of Child Labor” (Irvine, CA: University of California, October 1999) 18-19.

⁷ Christiaan Grootaert, “Modelling the Determinants of Child Labor” in C. Grootaert and H. A. Patrinos, eds., The Policy Analysis of Child Labor: A Comparative Study (New York: St. Martin’s Press, 1999) 19 [hereinafter “Determinants of Child Labor”].

⁸ This factor was found to be particularly important amongst very poor households in Cote d’Ivoire. See Christiaan Grootaert, “Child Labor in Côte d’Ivoire” in Policy Analysis of Child Labor at 27.

⁹ “Educational Composition of Labor Force” at 141-159.

¹⁰ C. Grootaert and R. Kanbur, “Child Labour: An Economic Perspective,” International Labour Review 134 (2) (1995) 193 [hereinafter “Child Labour: An Economic Perspective”].

¹¹ “Determinants of Child Labor” at 19.

¹² Dov Chernichovsky, “Socioeconomic and Demographic Aspects of School Enrollment and Attendance in Rural Botswana,” Economic Development and Cultural Change 33 (2) (1985) 328 [hereinafter “School Enrollment and Attendance in Rural Botswana”].

¹³ “School Enrollment and Attendance in Rural Botswana” at 328, as cited in H. A. Patrinos and G. Psacharopoulos, “Family size, schooling and child labor in Peru—An empirical analysis,” Journal of Population Economics (Springer- Verlag, 1997) 10: 387-405 [hereinafter “Family size in Peru”]. Patrinos and Psacharopoulos employ the term “specialization” to refer to the finding described by Chernichovsky (p.328) in his piece on rural Botswana.

¹⁴ “Hard-Headed Look at Labor Standards,” in International Labor Standards and Global Economic Integration Symposium at 31; and Jagdish Bhagwati, “A View from Academia,” in International Labor Standards and Global Economic Integration Symposium at 59.

¹⁵ Kenneth A. Swinnerton, “An Essay on Economic Efficiency and Core Labor Standards,” World Economy 20 (1) (1997)

¹⁶ Kimberly Cartwright, “Child Labor in Colombia,” in Policy Analysis of Child Labor at 78.

¹⁷ K. Cartwright and H. A. Patrinos, “Child Labor in Urban Bolivia”, in Policy Analysis of Child Labor at 117 [hereinafter “Child Labor in Urban Bolivia”].

¹⁸ “Uganda’s Report and Position on Child Labor,” (Kampala: The Republic of Uganda, Jan. 1998) 35 [hereinafter “Uganda’s Report on Child Labor”].

¹⁹ F. Angeles-Bautista and J. Arriola, To Learn and To Earn: Education and Child Labor in the Philippines (Manila: ILO/ IPEC, 1995) 14.

²⁰ J. Hallak and F. Caillods, ed. Educational Planning: The International Dimension (Geneva: UNESCO, 1995) 146.

²¹ “Child Labor and Schooling in Ghana” at 27.

²² Bjørne Grimsrud and Liv Jorunn Stokke, Child Labour in Africa: Poverty or Institutional Failure? The Cases of Egypt and Zimabawe (Norway: Fafo Institute for Applied Science, 1997) 16 [hereinafter The Cases of Egypt and Zimabawe].

²³ “Child Labor in Urban Bolivia” at 123.

²⁴ L. M. Sachikonye, Child Labour in Hazardous Employment: The Case of Zimbabwe, Consultancy Report Series no. 18 (Harare: Zimbabwe Institute of Development Studies, 1991) as cited in The Cases of Egypt and Zimabawe at 15.

²⁵ Rene Loewenson, Child labour in Zimbabwe (Harare: Study Report, 1992) as cited in The Cases of Egypt and Zimabawe at 15.

²⁶ “Family size in Peru” at 387-405 as cited in Kaushik Basu, “Child Labor: Cause, Consequence, and Cure with Remarks on International Labor Standards,” Journal of Economic Literature 37 (September 1999) 1093 [hereinafter “Child Labor: Cause, Consequence, and Cure”].

²⁷ “Child Labor: Cause, Consequence, and Cure” at 1093.

²⁸ “Child Labour: An Economic Perspective” at 194 .

²⁹ In instances where, for example, entire families work the same land, a poor harvest may mean that all family members’ income is disrupted. Child labor, while it may be necessitated for other reasons, is probably not a means of diversifying risk in examples such as these.

³⁰ “Determinants of Child Labor” at 6.

³¹ Policy Analysis of Child Labor at 6, 20.

³² H. Jacoby and E. Soufias, “Risk, Financial Markets and Human Capital in a Developing Country”, Mimeo, World Bank Policy Research Department (Washington, D.C.: World Bank, 1994) as cited in “Child Labour: An Economic Perspective” at 194.

³³ Christiaan Grootaert, “Child Labor in Cote d’Ivoire” in Policy Analysis of Child Labor at 25-27 [hereinafter “Child Labor in Cote d’Ivoire”].

³⁴ See Eddy Lee, The Asian Financial Crisis: the Challenge for Social Policy (Geneva: International Labor Office, 1998)

³⁵ A study in Venezuela found that parental education has a significant and positive effect on children’s completion of years of schooling; see “Educational Attainment in Venezuela” at 292. A study in Paraguay found that parental education has a negative influence on whether children work; see “Education and Earnings in Paraguay” at 57-68.

³⁶ “Child Labor in Cote d’Ivoire” at 45. In urban Cote d’Ivoire, each additional year of a father’s education raises the probability that a child only goes to school or combines work and school; a mother’s education has a significant effect on the probability of a child combining work and school, but also reduces the number of hours worked by a child. In rural areas, a father’s education raises the probability of a child combining work and school and also raises the number of hours that a child is likely to work. A mother’s education raises the probability of a child attending school and not working and of a child combining work and school. Ibid. at 44-45, 50-51.

³⁷ Kimberly Cartwright, “Child Labor in Colombia,” in Policy Analysis of Child Labor at 96. Cartwright finds that children are less likely to work the more educated either parent.

³⁸ “Child Labor in Urban Bolivia” at 122. Cartwright and Patrinos find that the more educated a child’s mother, the lower the probability that the child works.

³⁹ C. Sakellariou and A. Lall, “Child Labor in the Philippines,” in Policy Analysis of Child Labor at 145 [hereinafter “Child Labor in the Philippines”]. Sakellaiou and Lall find that in urban areas the less educated a head of household, the more likely children in that household will work.

⁴⁰ For a more complete look at the issue of educational access, see By the Sweat and Toil of Children, Volume V: Efforts to Eliminate Child Labor (Washington,D.C.: U.S. Department of Labor, 1998) Chapter IV.

⁴¹ State of the Worlds Children 1999 (New York: UNICEF, 1999) 33.

⁴² “Child Labor in Cote d’Ivoire” at 57.

⁴³ “Educational Outcomes in Rural Ghana” at 303.

⁴⁴ The Cases of Egypt and Zimabawe at 39.

⁴⁵ The tabulations in Appendix B illustrate a general pattern of higher economic activity among rural children.

⁴⁶ Kiran Bhatty, et. al., “Class Struggle,” India Today 22 (1997) 69-73 as cited in “Child Labor: Cause, Consequence, and Cure” at 1089.

⁴⁷ “Child Labor in Cote d’Ivoire” at 57.

⁴⁸ “Child Labor in the Philippines” at 151.

⁴⁹ State of the Worlds Children 1999 (New York: UNICEF, 1999) 9.

⁵⁰ “Education: Sector Strategy Paper” 1996-1998 (Bangalore: ACTIONAID, 1996) 16.

⁵¹ “Uganda’s Report on Child Labor” at 39.

⁵² E. A. Hanushek and V. Lavy, “Dropping Out of School: Further Evidence on the Role of Schooling Quality in Developing Countries,” University of Rochester Center for Economic Research, Working Paper No. 345 (Rochester: March 1993) 28.

⁵³ The Cases of Egypt and Zimabawe at 31.

⁵⁴ U.S. Embassy-Nairobi, unclassified cable No. 005357, May 05, 1999.

⁵⁵ “Uganda’s Report on Child Labor” at 39.

⁵⁶ The ILO estimates that worldwide about 27 percent of boys five to 14 years old are economically active, compared to 22.3 percent of girls. “Statistics on Child Labor in Brief” (http://www.ilo.org/public/english/120stat/actrep/ childhaz.htm).

⁵⁷ This is less likely to be the case in higher income families and in households with more highly educated mothers since such households are less likely to need to rely on girls to assume responsibility for domestic work. “Child Labour: An Economic Perspective” at 192; see also C.O.N. Moser, “Adjustment from Below: Low-Income Women, Time and Triple Role in Guayaquil, Ecuador,” in H. Afshar and C. Dennis (eds), Women and Adjustment Policies in the Third World (New York: St. Martin’s Press, 1992) 99. See also C. Grootaert and H. A. Patrinos, “The Policy Analysis of Child Labor” in Policy Analysis of Child Labor at 5 [hereinafter “Policy Analysis”].

⁵⁸ S. Cochrane, V. Kozel, and H. Alderman, “Household Consequences of High Fertility in Pakistan,” World Bank Discussion Paper No. 111 (Washington, D.C.: World Bank, 1990) as cited in “Policy Analysis” at 4.

⁵⁹ Priorities and Strategies for Education: A World Bank Review (Directions in Development) (Washington, D.C.: World Bank, 1995) as cited in “Policy Analysis” in Policy Analysis of Child Labor at 5.

⁶⁰ “Child Labor in Cote d’Ivoire” at 47. See also Policy Analysis of Child Labor at 5.

⁶¹ M. R. Rosenweig and R. E. Evenson, “Fertility, schooling, and the economic contribution of children in India: An economic analysis,” Economerica 45 (1977) 1065-1079, as cited in E. Skoufias, “Labor Market Opportunities and Intrafamily Time Allocation in Rural Households in South Asia,” Journal of Development Economics 40 (1993) 293.

⁶² Policy Analysis of Child Labor at 5.

⁶³ Santosh Mehrotra, Education For All: Policy Lessons From High-Achieving Countries, UNICEF Staff Working Papers, Evaluation, Policy and Planning Series (New York: UNICEF, 1998) 11.

⁶⁴ A. Bequele and J. Boyden, eds., Combatting Child Labor (Geneva: ILO, 1988) as cited in Cartwright, Kimberly, “Child Labor in Colombia,”as cited in F. Siddiqi and H. A. Patrinos, “Child Labor: Issues, Causes and Interventions,” Human Resources Development and Operations Policy Working Paper #56 ( June 1995) 8.

⁶⁵ G. Mohan Kumar, “Child Labour in Mosaic Chip Industries and Limestone Kilns in Kurnool District of Andhra Pradesh” in R. Anker, S. Barge, S. Rajagopal, and M. P. Joseph, eds., Economics of Child Labour in Hazardous Industries of India (Baroda, India: Centre for Operations Research and Training, 1998) 187 [hereinafter Child Labour in Hazardous Industries of India].

⁶⁶ Myron Weiner, The Child and the State in India (Princeton: Princeton University Press, 1991) 5.

⁶⁷ A similar phenomenon exists in Nepal, where children from the lowest social class–the so-called “untouchables”–are denied the right to attend the same schools as upper-caste children. See State of the Rights of the Child in Nepal 1998: Country Report Released by CWIN (Kathmandu: Child Workers in Nepal Concerned Center, January 1998) 18.

⁶⁸ “Child Labor in Urban Bolivia” at 127.

⁶⁹ Ibid.

⁷⁰ State of the World’s Children 1999 (New York: UNICEF, 1999) 41.

⁷¹ Alec Fyfe, Child Labor (Oxford: Polity Press: 1989) 12, 28.

⁷² C. Turbay and E. Acuña, “Child Labor and Basic Education in Colombia” in M. Salazar, W. Glasinovich, eds., Child Work and Education: Five Case Studies from Latin America (Florence: UNICEF International Child Development Center, 1998) 41 [hereinafter Child Work and Education].

⁷³ D. Levison, R. Anker, S. Ashraf, and S. Barge “Is Child Labor Really Necessary in India’s Carpet Industry?” in Child Labour in Hazardous Industries of India at 99.

⁷⁴ Child Labour in Hazardous Industries of India at 20.

⁷⁵ Maruicio García-Moreno, “Child Work and Education in Ecuador” in Child Work and Education at 91.

⁷⁶ I. Rizzini, I. Rizzini, F. Borges, “Brazil: Children’s Strength is Not Their Work,”in Child Work and Education at 35.

⁷⁷ Correspondence from Bernd Balkenhol, Head, Social Finance Unit, Employment Sector of the International Labor Organization to U.S. Department of Labor Official (November 12, 1999) [document on file].

⁷⁸ Inter-American Development Bank, Facing Up to Inequality in Latin America (Baltimore: Johns Hopkins University Press, 1998) 167-168 [hereinafter Facing Up to Inequality].

⁷⁹ M. Woodhall, “Designing a Student Loan Program for a Developing Country: The Relevance of International Experience,” Economics of Education Review, 7 (1) 1 (1988) 153-61; and Lars Ljungvist, “Economic Underdevelopment: The Case for Missing Markets for Human Capital,” Journal of Development Economics, 40 (2) (April 1993) 220. See also, Priya Ranjan, “An Economic Analysis of Child Labor,” Economics Letters 64 (1999) 99-105.

⁸⁰ See Appendix A.

⁸¹ For a review of evidence from Latin America example, see Facing Up to Inequality at 166.

⁸² Ibid. at 164. See also Social Finance Unit Annual Report 1998: The Social Dimension of Poverty Alleviation, Employment and Social Integration (Geneva: International Labor Organization, 1998).

⁸³ See also Box III-1.

⁸⁴ By the Sweat and Toil of Children, Volume II: The Use of Child Labor in U.S. Agricultural Imports and Forced and Bonded Child Labor (Washington, D.C.: U.S. Department of Labor, 1995) 81-82.

⁸⁵ A. Bequele and J. Boyden, eds., Combatting Child Labor (Geneva: ILO, 1988) as cited in “Child Labour: An Economic Perspective” at 195.

⁸⁶ “Economics of Child Labor” at 416-419.

⁸⁷ D. Levison, R. Anker, S. Ashraf, and S. Barge, “Is Child Labour Really Necessary in India’s Carpet Industry?” in Child Labour in Hazardous Industries of India at 100 [hereinafter “Is Child Labor Really Necessary?”].

⁸⁸ “Child Labour: An Economic Perspective” at 195-196.

⁸⁹ The survey included data from the two districts in India (Uttar Pradesh, and Mirzapur and Sonbhadra) in which 80 percent of India’s carpets are produced. “Is Child Labour Really Necessary?” at 95, 108-115.

⁹⁰ S. Barge, R. Anker, S. Ashraf, and D. Levison, “Child Labour in Glass-Bangles Industry of Ferozabad—Uttar Pradesh: An Economic Analysis” in Child Labour in Hazardous Industries of India at 63-64.

⁹¹ The survey focused on certain sections of Surat city where the diamond processing industry is centered. Ranjana Kolhe Saradhi, “Economics of Replacing Child Labour in Diamond Industry of Surat” in Child Labour in Hazardous Industries of India at 82-83.

⁹² The survey focused on gem stone processing in the city of Jaipur. Nisha Lal, “Economics of Eliminating Child Labour in Gem Stone Industries” in Child Labour in Hazardous Industries of India at 159-161.

⁹³ G. Mohan Kumar, “Child Labour in Mosaic Chip Industries and Limestone Kilns in Kurnool District of Andhra Pradesh” in Child Labour in Hazardous Industries of India at 189.

⁹⁴ In 1995-1996, ILO/IPEC’s South-East Asia Office in Bangkok collaborated with the ILO’s Manila-based Multidisciplinary Advisory Team for South-East Asia and the Pacific (SEAPAT) on “A survey of child labour in South-East Asian manufacturing industries.” The survey included studies of the informal manufacturing sectors for garments, rattan furniture, footwear, and gemstone polishing. See Programme to combat Child Labor in the Footwear sector in South-East Asia (Indonesia, the Philippines and Thailand) (Phase II): Multi-bilateral Programme of Technical Cooperation (Geneva: ILO/IPEC, October 1993) 3.

⁹⁵ D. Galbi, “Child Labour and the Division of Labour” (Cambridge, U.K.: Kings’s College Centre for History and Economics, 1994) as cited in “Child Labour: An Economic Perspective” at 196.

⁹⁶ Mark R. Rosenzwieg, “Household and Non-Household Activities of Youths: Issues of Modelling, Data and Estimation Strategies,” in G. Rodgers and G. Standing, eds., Child Work, Poverty and Underdevelopment (Geneva: International Labor Office, 1981) as cited in “Child Labour: An Economic Perspective” at 196.

⁹⁷ Victor Levy, “Cropping Pattern, Mechanization, Child Labour, and Fertility Behavior in a Farming Economy: Rural Egypt,” Economic Development and Cultural Change 33(4) (1985) as cited in “Child Labour: An Economic Perspective” at 195-196.

⁹⁸ D. S. DeGraff, R. E. Bilsborrow, and A. N. Herrin, “The Implications of High Fertility for Children’s Time Use in the Philippines,” in C. B. Lloyd, ed., Fertility, Family Size and Structure – Consequences for Families and Children (New York: The Population Council, 1993) as cited in “Child Labour: An Economic Perspective” at 196.

⁹⁹ Maria Cristina Salazar, “Child Labour in Colombia: Bogotá’s Quarries and Brickyards,” in Assefa Bequele and Jo Boyden, eds., Combatting Child Labour (Geneva: International Labor Office, 1988) 51-52, as cited in “Child Labour: An Economic Perspective” at 196.