Understanding the links between education and smoking

Educational gradients in smoking are one of the deadliest examples of social disparities in health. But is this association between education and smoking casual? If it is, and we understood which aspects of schooling caused individuals not to smoke, then educational policy could have massive health dividends. In contrast, if the relationship between education and smoking is non-causal, then the observed gradients are instead explained by other characteristics that predict both statuses, making the disparities more difficult to address. This project has three goals: (1) to extend the theoretical discussion on the relationship between education and health to include the important life course links between characteristics in adolescence and outcomes in adulthood; (2) to estimate an empirical model of the relationship between education and smoking in adulthood using measures that account for the appropriate timing of the theoretically relevant mechanisms across the life course; and (3) to formulate a model of smoking and education in adulthood that considers the bundling of these characteristics and how their joint distribution relates to personal, family, school, and smoking-related characteristics in adolescence, net of future expectations. The results show that the relationship between education and smoking in adulthood is well-established by the early adolescence.