Child labour measurement: Whom should we ask?

The paper asks whether the type of respondent changes child labour statistics, using data from two sites in Tanzania. It finds that who is asked matters a great deal. Child self-reports give higher child labour prevalence than reports from proxies such as adults answering on the child's behalf. The gap shrinks as children get older and grows when the proxy opposes child labour. Funding decisions rely on these statistics, so measurement accuracy is important.

Key points

• Prevalence of child labour is about 35 to 65 per cent higher when child self-reports are used instead of proxy reports.
• This discrepancy affects 14 to 31 per cent of the sample, depending on the indicator.
• Discrepancies decrease as the child ages.
• Discrepancies increase when proxy attitudes show opposition to child labour.
• Funding decisions to support vulnerable children are tied to child labour statistics.

Limitations

Only the abstract is available, so the survey design, sample sizes, indicators and methods are not described. The data come from two sites in Tanzania, and the abstract does not say whether the findings apply to other places or which report is more accurate.