UC Irvine

Background: Divergence of male and female life expectancy is a well-documented phenomenon. Comovement is a heretofore-neglected aspect of changes in male and female mortality. Objective: We develop a new framework for life expectancy sex differentials in time series, using co-movement/anti-movement and convergence/divergence. METHODS We apply this framework to the Human Mortality Database (HMD), assessing co-movement between male and female life expectancy with the nonparametric test of Goodman and Grunfeld (1961). Results: For every country in the HMD (except three with short spans of data), male and female mortality statistically co-move. This applies even in cases, including ones such as Russia that are well-discussed in the literature, that show extreme divergence between the sexes. The results are reasonably robust to subsetting with a 25-year time-window for all countries. Conclusions: Male and female life expectancy co-move even when the life expectancy sex differential increases. The sex divergence in life expectancy needs to be (re-)considered in light of the fact that male and female life expectancy usually co-move, reflecting overall societal factors.