In many empirical studies mortality differences between socioeconomic groups (SES) decrease in the
higher age groups. However, the mechanism behind this convergence is unknown. This study presents
empirical evidence and possible explanations. Danish register data of all men in Denmark above age
58 between 1980 and 2002 (n=938.427) and event history analysis is used to study mortality
differences between income groups, controlled for eight other variables. Interaction models with age
or health status are used to describe the change of SES mortality differences with age. Mortality
differences in Denmark are very large. The upper 75 percent of the income distribution have very
similar mortality levels, but have approximately only 35 percent of the mortality risk of the poorest 10
percent. Mortality differentials are stable across age groups (controlled for health) but they converge
completely when health is deteriorating. This study shows that instead of “age as leveler” it is “illness
as leveler”. The finding that SES only has a very small impact on the transition from poor health to
death shows that SES mortality differences do not exist because ill people with low SES have poor
access to intensive or expensive medical care. It rather suggests that SES differences in mortality
originate in the period of prevention and early treatment. This is also the period where policy measures against health inequality are most promising.
The fulltext (PDF) was removed on 28.07.2009 and re-uploaded on 30.01.2015.
