LAUSR

Using Swedish data on a cohort born in 1953, interviewed in 1966 (age 13), and followed with register data until 2018 (age 65), this study shows that one more year of schooling predicts a 17% lower risk of early mortality. Addressing concerns of potential selection bias, the mortality inequality by educational attainment persists when extensive controls are included in the regression. Adding information on background health, gender, socioeconomic variables, as well as adolescents' early educational plans, cognitive ability, and time preferences, only result in a 2-percentage point change in the mortality risk by years of education. Even when adolescents' applications to upper-secondary school and years 6 and 9 grades are controlled for, completion of upper-secondary and university education remain strong predictors of future health. Yet, the study also finds that the measure of future health matters for the stability of the results.