Addressing social desirability bias in child labor measurement: An application to cocoa farms in Côte d’Ivoire
This study, published in the Bordeaux Economics Working Papers, compares two methods – direct questioning and list experiments – to estimate the prevalence of child labour in a sample of 4,458 certified cocoa farms in Côte d’Ivoire.
By using an indirect questioning method, the researchers estimated the proportion of certified cocoa farmers who use child labour to perform hazardous work on their farms.
The study found that social desirability appears to play an important role. The indirect questioning method, using list experiments, gave much higher estimates of child labour prevalence – double the rate provided by farmers interviewed directly.
The results of the list experiment suggest that 24% of farmers received the help of a child under the age of 16 on their farm for harvesting or the breaking of cocoa pods during the past 12 months, 21% of farmers received help to prepare their farm, while 25% employed and paid a child to perform any task on their farm.
The prevalence of child labour was also found to be higher in remote communities with limited school opportunities, lower access to adult labour, and weaker law enforcement.