Child Labor Data: What the 138 Million Count Actually Captures

The United Nations counted 138 million children in child labor in 2024, about 8% of children aged 5 to 17 worldwide. The figure is often quoted as a single headline. The definition behind it is not uniform, spanning an 11-year-old feeding animals on a family farm and a child working 43 hours a week in a factory. For investors mapping supply chain exposure, the total matters less than the composition.

How the UN Defines Child Labor

Measurement rules were settled in 2008 by the International Conference of Labour Statisticians, building on conventions that go back to the ILO's first conference in 1919. The rules rest on age and hours, with the nature of the job deciding what counts as hazardous. Any work by a child under 12 is child labor, even light and non-hazardous tasks. For ages 12 to 14, the threshold is 14 hours a week. For ages 15 to 17, it is 43 hours. Hazardous work, meaning dangerous industries or occupations, or any work at 43 hours or more per week, is counted at any age.

Under those rules, 54 million children are in hazardous work, about 3% of all children. The remaining 84 million are in non-hazardous jobs that still qualify, mostly because the children are too young or working too many hours. The harm in those cases is not the work itself. It displaces schooling and rest.

Close to half of all child labor involves non-hazardous work by children under 12. Around 6 in 10 children in child labor work on farms or in other agriculture, most of it on smallholder plots run by their families. An 11-year-old feeding animals for a couple of hours each morning and afternoon counts as child labor even when the work is light. The same child working 43 hours or more across the week moves into the hazardous category, counted alongside children in construction and mining.

Where the Estimates Come Up Short

Domestic chores in a child's own home, such as cooking and childcare, are excluded. Only economic activity counts, meaning employment or producing goods. A child spending the same hours caring for siblings at home is not counted, while a child feeding animals on a farm is. That exclusion skews the figures toward boys, because girls do more of the household work that falls outside the definition. The UN's broader measure, which adds domestic work of 21 hours or more per week, reverses the gender gap.

Some of the worst forms of child labor, including trafficking and forced labor, are not captured at all. They fall outside what household surveys can measure, a limitation rather than a conceptual choice. The ILO tracks modern slavery in a separate estimate with different data sources, so the two sets of numbers do not line up.

Under the UN's measurement framework, hazardous work is the category the international conventions target most directly, and it covers 54 million children. The concentration in agriculture matters for a different reason: most of that work happens on family plots, largely outside formal supply chains. The measurement gaps add a separate layer of risk. Hardware makers, Apple (AAPL) among them, have run supplier audits on labor conditions for years.

The world set a target under the Sustainable Development Goals to end child labor in all its forms by 2025. The deadline has passed. The 2024 estimate put the headcount at 138 million, or 8% of children aged 5 to 17.