What Can Lifespan Variation Tell Us about the Future of Mortality? The Curious Case of High-School Educated Americans

Isaac Sasson, University of Texas at Austin

The educational gradient in U.S. adult mortality is well documented, but focuses almost exclusively on differences in life expectancy (“mean age at death”). Other dimensions of lifespan inequality, such as lifespan variation, are often neglected. From an individual standpoint, lifespan variation represents higher uncertainty in the time of death; from a population perspective, it represents higher group heterogeneity. Using data from the National Vital Statistics System, 1990-2010, this study decomposes educational differences across the entire age-at-death distribution into their constituent parts (means and variances). Findings reveal that although differences in means generally dominate over differences in variances, the latter are becoming increasingly important among low and high-school educated Americans in most race-sex groups. For some, differences in lifespan variation even surpass differences in life expectancy in importance. A multi-dimensional approach to lifespan inequality sheds light on competing mortality scenarios, with implications for understanding, modeling, and predicting the future of mortality.