Earlier studies of variation in mortality between population groups have usually not considered the combined importance of two or more sociodemographic factors. Thus, they have provided an inadequate description of how large the variation actually is. In this register-based study of Norwegians aged 50–89 in 1975–2008, a richer picture is painted by using a variable that combines marital status and own and spouse’s education. There is a large difference in mortality between the extreme categories, which are the divorced or never-married with low education and the married with tertiary education whose spouse also has tertiary education. In 2005–08, the age-standardized death probabilities differed by a factor of more than three, and the difference in life expectancy was 9.4 years for men and 8.5 for women. These ratios and differences have increased over the 34-year period, and the variation has also increased according to measures that take all categories of the combined variable into account. The increasing variation is partly a result of rising mortality among never-married people with low education. This large and growing social variation in mortality may be seen as a motivation for intensified discussions of how social health inequality can be mitigated.
Published in a peer-reviewed journal of the Population Council. Øystein Kravdal is Professor of Demography, University of Oslo, and Centre for Fertility and Health, Norwegian Institute of Public Health.