Olof Östergren

Aims: Social inequalities in mortality persist or even increase in high-income countries. Most evidence is based on a period approach to measuring mortality – that is, data from individuals born decades apart. A cohort approach, however, provides complementary insights using data from individuals who grow up and age under similar social and institutional arrangements. This study compares income inequalities in cohort life expectancy in two Swedish cohorts, one born before and one born after the expansion of the welfare state. Methods: Data on individuals born in Sweden in 1922–1926 and 1951–1955 were obtained from total population registries. These data were linked to individual disposable income from 1970 and 1999 and mortality between 50 and 61 years of age in 1972–1987 and 2001–2016, respectively. We calculated cohort temporary life expectancies in the two cohorts by income and gender. Results: Life expectancy, income, and income inequalities in life expectancy increased between the two cohorts, for both men and women. Women born in 1922–1926 had modest income differences in life expectancy, but pronounced differences emerged in the cohort born in 1951–1955. Men with low incomes born in 1951–1955 had roughly similar life expectancy as those with low incomes born in 1922–1926. Conclusions: Compared with a period approach to life expectancy trends, the cohort approach highlights the stagnation of mortality at the lowest income groups for men and the rapid emergence of a mortality gradient for women. Future research on health inequalities in welfare states should consider underlying factors both from a cohort and period perspective.