Population Research and Policy Review

, Volume 32, Issue 6, pp 893-917

First online:

The Nonlinear Relationship Between Education and Mortality: An Examination of Cohort, Race/Ethnic, and Gender Differences

  • Bethany G. EverettAffiliated withDepartment of Sociology, University of Illinois-Chicago Email author 
  • , David H. RehkopfAffiliated withDivision of General Medical Disciplines, Stanford University School of Medicine
  • , Richard G. RogersAffiliated withDepartment of Sociology and Population Program, Institute of Behavioral Science, University of Colorado-Boulder

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


Researchers investigating the relationship between education and mortality in industrialized countries have consistently shown that higher levels of education are associated with decreased mortality risk. The shape of the education–mortality relationship and how it varies by demographic group have been examined less frequently. Using the U.S. National Health Interview Survey-Linked Mortality Files, which link the 1986 through 2004 NHIS to the National Death Index through 2006, we examine the shape of the education–mortality curve by cohort, race/ethnicity, and gender. Whereas traditional regression models assume a constrained functional form for the dependence of education and mortality, in most cases semiparametric models allow us to more accurately describe how the association varies by cohort, both between and within race/ethnic and gender subpopulations. Notably, we find significant changes over time in both the shape and the magnitude of the education–mortality gradient across cohorts of women and white men, but little change among younger cohorts of black men. Such insights into demographic patterns in education and mortality can ultimately help increase life expectancies.


Education Mortality Race/ethnicity Gender Semiparametric modeling NHIS