Determinants of Life Expectancies in U.S. Counties

There is a large variability in both female and male life expectancies by county in the U.S. We develop and test a model of the impact of demographic, economic, educational, social, and geographic factors on mean life expectancy by county for males and females born in 1990 . We find that the percentage of population on rural farms, the percentage of married households, the level of education, the percentage speaking a language other than English at home, the percentage foreign-born, and county elevation have significant positive effects on life expectancy for both males and females; while the percentage of population below the poverty level, violent crime rate, population density, unemployment rate, and latitude have significant negative effects. Income has a nonlinear effect on life expectancy, but a negative impact on average female life expectancy. The percentage of the population reporting Northern European ancestry increases mean county life expectancy, while the percentage black or Native American reduces life expectancy. Finally, we identify state effects on mean county life expectancies, including the Mason-Dixon Effect, the Cowboy Effect, and-for females-the Rust Belt Effect