Parental love is not blind: Identifying selection into early school start

Do parents take into account their children’s ability when deciding on their education? If so, are parents’ perceptions accurate? We study this by analyzing a key educational decision. Parents choose whether their children start elementary school one year early. Do they select high-ability children to start early? We propose a novel methodology to identify the sign and strength of selection into early starting. We find robust evidence of positive selection. Had they started regularly, early starters would have obtained test scores 0.2 standard deviations higher than the average student. We also estimate the effect of early starting on test scores and find that the effect is heterogeneous.