Exploiting a rich panel data child survey merged with administrative records along with a pseudo-experiment generating variation in the take-up of pre-school across municipalities, we provide evidence of the effects on non-cognitive child outcomes of participating in large scale publicly...