The Impact of Smoking And Other Non-Biological Factors on Sex Differences in Life Expectancy: An International Comparison

The aim of this paper is to investigate more deeply the role of smoking for sex differences in life expectancy in comparison to other non-biological factors. Based on the findings of previous studies we expect that populations differ mainly with respect to the time location inside the smoking epidemic model and the absolute number of years smoking contributes to the sex gap. To test these hypotheses we decompose the sex differences between 1955 and 2009 into fractions caused by biological factors (estimated on the basis of differences in life expectancy between female and male Catholic order members), smoking, and other non-biological factors for 53 industrialized countries. We find that smoking can indeed be seen as the main driver of the trend in sex differences in life expectancy for most populations, giving further support to the importance of the smoking epidemic model. However, our results reveal that the common view that smoking is also responsible for most of the sex difference itself does not hold in general. There are almost as many populations with a higher contribution of other non-biological factors as populations in which smoking is the main factor behind the sex gap.