If you ever tell people child labor bans are bad, expect some side-eyes (that wording is only good for clickbait).
And to be clear, children working is bad. Hard labor sucks for everyone, but especially children, who are considered society’s most vulnerable. Hence why we in the developed world like to protest companies found employing ten year olds, like Nike.
Also why the 1930s movement to end child labor in America is remembered as a heroic fight, immortalized in many a YouTube video and history textbook. And why, in the 1970s, the UN’s International Labor Organization passed ILO Convention 138, the Minimum Wage Convention, calling for an end to the labor of children under 15.
All that said, you shouldn’t ban it.
What happens when you ban child labor?
The issue with child labor bans isn’t that child labor is good; it’s that banning child labor doesn’t fix child labor. The opposite, in fact. When India passed the Child Labor (Prohibition and Regulation) Act of 1986, child worked more, wages went down, children attended school slightly less often, and caloric intake decreased.
On its face, this doesn’t make sense. Why would a ban encourage child labor? Consider the reason a child works in a poor family. Most parents don’t want their kids to work in the mines or the fields, but the adults’ income alone isn’t enough to pay for food and shelter. The parents aren’t lazy people using their kids to make extra income, they’re people using their kids to survive. As Kaushik Basu and Pham Hong Van, some of the most prominent economists to study the topic, phrased it in a landmark 1998 paper:
[Children working in low-income countries] reflects not a difference in the attitude of the parents but the problem of stark poverty where the parents are compelled to send the children to work for reasons of survival. Even in England, which witnessed some of the worst excesses of child labor in the late eighteenth and early nineteenth centuries, a parliamentary report noted that “parents were desperately unhappy about the situations their children were in but could do nothing about it. The social system allowed them no choice.”
When the choice between sending a child to work and sending them to school is the choice between satiation and hunger, it’s not a surprise that many choose the former, even with a ban.
Still, that only explains why child labor endured. After India banned working under 15, the probability a child under 15 worked rose 7.8%, or 2.6 percentage points (relative to a child over 14; the overall rate was decreasing among all age groups throughout this period as parental incomes increased). Child labor didn’t just endure, it grew.
It makes sense if you think about the employer’s perspective. India enforced their ban by fining companies and jailing managers caught employing children. This effectively raises the cost of employing a child: say you think there’s a 10% chance you’ll be caught and will pay the 10,000-rupee fine. If you employee 50 children, that means you expect each of them to cost you 20 rupees more than they did pre-ban. The substitution effect dictates that, since the price of child labor grew but the price of adult labor didn’t, employers will try to employ more adults. They correspondingly decrease their wages for children.
Which is a problem, if you needed those wages to feed your family. With lower wages, children need to work more in order to earn the same amount of money. And since the children want more hours, the supply of labor increases, pushing wages for everyone — adult or child — down further. The result is children working more, earning less, (in some cases) attending school less, and often still failing to earn enough, hence why they consume fewer calories. In a word, it’s bad.
I should mention that this is only one example, albeit the most prominent one. After Brazil raised its minimum age for employment from 14 to 16 in 1998, the proportion of urban boys working dropped 35%, with other children being unaffected. Most of these boys had been both working and studying prior to the ban and transitioned to solely studying. There didn’t appear to be long-term effects on their percentage of employment, formal occupation, wage per hour, and undergrad enrollment or completion.
The lack of downsides might be explained differences between 1998 Brazil and 1986 India. First, going from a 14 minimum age for employment to 16 is different from going from no minimum age to 14. In the Brazilian study, the people affected are those who would’ve started working at 14 or 15 but had to hold off until 16 under the new law. In the Indian study, the people affected are those who would’ve started working before puberty — a group in more desperate need of money. 1998 Brazil was a lot more developed than 1986 India. As a rough comparison, Brazil’s 1998 GDP per capita was about 5,050 current USD, about 15 times India’s 1986 figure of 320.
More food for thought: this theoretic model based on data from 1880 found that banning child labor would lead to overall welfare increases because parents would work more.
So, how do you discourage child labor?
Not easily. In the early 1990s, child workers were found producing clothing for Wal-Mart, and Congress hoped to deter child labor by banning child-produced imports. The effort culminate in the proposal of the US Child Labor Deterrence Act, better known as the Harkin Bill after its sponsoring congressman. It never passed — but it became popular, which meant they succeeded in reducing child labor! But the kids didn’t go back to school; they ended up on the streets as prostitutes or street hustlers.
Child labor is more of a symptom of poverty than a cause. Developed countries have low child labor rates because families are rich enough not to need their kids to work, not because it’s illegal.
Still, there are some effective options. In 1997, Mexico launched a system of conditional cash transfers (under the broader Progresa assistance program) that paid poor families if their children attended school. The value of the transfers was equivalent to about 2/3 of the income the child would’ve earned working, and was enough to convince about 10% of families affected to begin sending their children to school. It amounted to an average of 0.66 additional years of additional schooling, from 6.88 years before the program.
That’s despite the fact the reduced labor supply meant wages for children increased. This would be a success if it were only about reducing child abuse in the workforce or alleviating poverty. Long-term, though, it also raised later-life incomes, increased geographic mobility, and postponed family formation, all key indicators of upward mobility.
Education remains one of the most powerful mechanisms for social progress, and labor necessarily an obstacle to it — to say less of the abuses which can occur in the workforce. But children in these scenarios are stuck between a rock and a hard place: the alternative to work is not first-world idyll but something worse. When we design policies to move beyond child labor, we have to do so more thoughtfully. To treat it as the worst thing that could happen to a child is to lose nuance. Labor is often the best of bad options; good development policy considers that. Real world morality is not black and white.