Advocates of a universal child care system offer a two-fold argument: Child care facilitates children's long-run development, and levels the playing field by benefiting in particular disadvantaged children. Therefore, a critical element in evaluating universal child care systems is to measure the impact on child development in a way that allows the effects to vary systematically over the outcome distribution. Using non-linear DD methods, we investigate how the introduction of large-scale, publicly subsidized child care in Norway affected the earnings distribution of exposed children as adults. We find that mean impacts miss a lot: While child care had a small and insignificant mean impact, effects were positive over the bulk of the earnings distribution, and sizable below the median. This is an important observation since previous empirical studies of universal child care have focused on mean impacts. We further demonstrate that the essential features of our empirical findings could not have been revealed using mean impact analysis on typically defined subgroups. This is because the intragroup variation in the child care effects is relatively large compared to the intergroup variation in mean impacts.