# 12-070/4 (2012-07-17; 2013-10-30)

Jan S. Cramer, University of Amsterdam
Cognitive ability, mortality, socio-economic status, proportional hazards
JEL codes:
C21, I14

The initial purpose of this study was to establish the effect of childhood conditions on longevity from the Brabant data set. This data set combines information at ages 12, 43, 53 and mortality between 53 and 71 for a sample of some 3000 individuals born around 1940 in the Dutch province of North Brabant. Proportional hazard analysis confirms the known association of early intelligence or cognitive ability with longevity, with a standardized hazard ratio of .80; this is the only significant childhood influence. Among men, the effect of some elements of adult socio-economic status can also be ascertained: education, income and wealth are each found to contribute about as much to a longer life as intelligence. The joint effect of all four variables is dominated by childhood intelligence and adult wealth at the expense of education and income.