Education, cognitive ability and cause-specific mortality: a structural approach

Govert Bijwaard, Netherlands Interdisciplinary Demographic Institute (NIDI)
Mikko Myrskylä, Max-Planck Institute for Demographic Research and London School of Economics and Political Science (LSE)
Finn Rasmussen, Karolinska Institutet
Per Tynelius, Karolinska Institutet

Many studies show large differences in cause-specific hazard rates by education. These studies ignore that cause-specific hazard rates are interdependent and, more importantly, that educational attainment and cause-specific mortality may both depend on (latent) cognitive ability. We focus on the months lost due to a specific cause of death, a quantify with a more natural interpretation which avoids the interdependence. We account for possible confounding of cognitive ability by using a structural framework that explicitly models the correlation. We derive the educational gains, in months lost from age 18 till age 63, that account for both observed individual characteristics and for unobserved cognitive ability. We also derive the selection effects for each cause of death, and quantify the contribution of observed characteristics and unobserved cognitive ability in the selection effects. We use the Swedish Military Conscription Data (1951-1960), linked to administrative Swedish registers followed till the end of 2013. Our empirical findings reveal that the largest educational gains can be achieved for the lowest education group in the reduction of external causes of death. The educational gains in cardiovascular mortality is rather small, mainly due to large selection effects.