Previous findings on (fleeting) relative age effects in school suggest that, given innate ability, too few younger and too many older students attend academic tracks. Using a regression discontinuity design around school-specific admission thresholds, we estimate the cognitive and non-cognitive...