Measurement of child labor is critical for our understanding of its determinants and conditions, and for the design of social protection programs and policy. This study reports the findings of three survey design experiments implemented across Fairtrade coffee households in rural Ethiopia in three different agricultural seasons.

The study aimed to answer two questions: How does the timingof data collection affect child labour rates? And how do child labour rates differ, depending on who is asked?

The findings show that both the timing and respondent make a difference. Child labor participation varied substantially, depending on the season. Child labour rates ranged from 45 to 76 percent. They were highest during the harvest season, when households rely heavily on their children’s work. When choosing a respondent, gender matters. The work of girls in agricultural settings was systematically underreported by others, such as the head of household, compared to the child’s reports. In contrast, no reporting differences were found for boys, when a male household member responded on their behalf. In answer to other questions on schooling and household chores, there were no clear differences between reports of children and adults.

Written by by Jose Galdo, Ana C. Dammert and Degnet Abebaw, this study is part of the Growth and Labour Markets in Low Income Countries Programme, funded by IZA and the UK Government Department for International Development (DFID).

child labor measurement cover

Child labor measurement in agricultural households