(Abstract with IA podcast summary) Many educational institutions and experts have raised the alarm on the observation that social media use is dangerous for young people’s well-being and mental health. However, existing reviews on this issue do not provide definite answers that address the problems of causality and heterogeneity in social media use. This paper selects, reviews and discusses empirical studies that more rigorously analyze causality in the field using large samples and objective data over long stretches of time, while overlooking the heterogeneity problem. These studies adopt the ‘natural experiment’ approach to study staggered access to social media across the territory. The conclusion drawn for findings across the US, the UK, Germany, Italy and Spain is that social media generally harms young people’s well-being and mental health. In discussing these studies, which belong to the economics literature, the present paper suggests a novel theoretical interpretation: social media use becomes harmful because it displaces beneficial activities aimed at achieving future and pro-social purposes which would make young people less vulnerable to addictive use of social media. Supporting evidence emerges from reviewing studies that pay attention to causality. The paper thus makes evident the need for research to integrate different methods and disciplines focused on such a complex and urgent issue.

Published and downloadable here, with IA podcast summary (3000 downoads in the first month)