The increasing number of people reaching older ages poses a threat to the pension systems of highly developed countries. In Denmark and Sweden, statutory retirement age is indexed to life expectancy to account for mortality improvements of their populations. However, mortality improvements are not equal across subgroups within a country. For example, mortality improvement has been slower for individuals of lower socioeconomic status and raising retirement age could disproportionally affect these individuals. The aim of this study is twofold: understand the implications of indexing retirement age for future inequalities among subgroups of two welfare state’s populations; and investigate how the different mortality trends after age 50 observed in Denmark and Sweden are reflected in their retirement ages.
We use Danish and Swedish registry data (1987-2019), to aggregate individuals aged 50 or more based on their demographic characteristics and socioeconomic status (SES), defined according to equivalised disposable family income. Mortality for the SES groups is forecasted by using two coherent versions of the Mode model and the Li-Lee model until 2040. We compute period life tables by year, sex, and SES for Denmark and Sweden to estimate, among other measures, the probability of surviving until retirement age and the age at which the remaining life expectancy equals 14.5. In Denmark it represents the age when one should retire according to the current pension reform.
Preliminary results show that indexing retirement age to life expectancy in Denmark will have larger impact for individuals with low SES. The difference between the probability of surviving to 65 and to the current (indexed) retirement age is higher for individuals with low SES. Even though the reform makes the pension system robust to population aging, our results suggest that those with low SES are going to be more (negatively) affected by the reform.