Relative to adults, adolescents make more welfare-decreasing decisions, especially in the presence of peers. The consequences of these decisions result in substantial individual and societal losses in terms of lives lost, injury, hospitalisation costs, and foregone opportunities.

In this paper, we used a laboratory experiment with 12- to 24-year-old participants to deconstruct the decision-making into three unique components: people’s tendency to take risks in pure gain scenarios, people’s tendency to take risks in pure loss scenarios, and loss aversion (people’s tendency to weigh losses more than gains of the same size). Participants were repeatedly asked to choose between a safe monetary outcome and a lottery. Each participant made the same set of choices in private and when observed by an age-matched peer which allowed us to measured how peer observation affects individual’s choices.

We found that older adolescents (18-24 years old) are more likely to take risks when observed. This happens independent of whether the consequences of their choices can only be gains, only losses, or mixed. We did not find evidence of adolescents becoming less sensitive to losses versus gains under observation. Also, for younger adolescents, neither their risk attitudes nor their loss aversion is affected by peer observation. The paper explores possible channels through which this behavioural change occurs.

While it would be tempting to think that in the presence of peers, adolescents stop paying attention to the potential negative consequences of their actions, our results show that this is not the case. Instead in the presence of peers, adolescents’ relative weighting of losses to gains (loss aversion) increases. This is good news for policy suggesting that appealing to loss aversion should be especially effective at reducing harmful adolescent behaviours committed in the presence of peers.