Luck or cheating? A field experiment on honesty with children

We run an experiment to study the relationship between honesty, age and self-control. We focus on children aged between 5 and 15 as the literature suggests that self-control develops within such age range. We ask each child to toss a fair coin in private and to record the outcome (white or black) on a paper sheet. We only reward children who report white. Although we are unable to tell whether each child was honest or not, we speculate about the proportion of reported white outcomes. Children report the prize-winning outcome at rates statistically above 50% but below 100%. Moreover, the probability of cheating is uniform across groups based on child's characteristics, in particular age. In a second treatment we explicitly tell children not to cheat. This request has a dampening effect on their tendency to over-report the prize-winning outcome, especially in girls. Furthermore, while this effect in boys is constant with age, in girls it tends to decrease with age.