Silvia Barcellos (USC)

“The Distributional Effects of Education on Health”


The literature on the effects of education on health has focused almost exclusively on mean impacts. This paper studies the distributional effects of education on health, investigating treatment effect heterogeneity. To overcome the endogeneity of education we exploit a natural experiment, the Raising of School Leaving Age Order of 1972 (ROSLA) which increased the minimum school-leaving age from 15 to 16 years in England, Wales and Scotland. Using a regression discontinuity design and objective health measures, we estimate that staying in school until 16 reduced average body size and increased average blood pressure approximately 35 years later. Importantly, these opposite-sign effects occur in different parts of the health distribution. While obesity rates dropped by 7.5 percentage points, the prevalence of hypertension did not increase. Changes in labor market outcomes, SES and diet are mechanisms driving our main findings. Taken together our results point to the importance of considering different dimensions of health and of moving beyond the analysis of mean impacts when investigating the role of education in the production of health.