Socioeconomic Status and Working Life Expectancy: A Swedish Cohort Study

Longer life expectancy and fertility decline raise several potential challenges for national economies and the security of welfare systems. Raising retirement ages, or ages of eligibility for state pension, is one strategy to offset rising costs. Yet, considerable subgroup variation exists in full retirement timing, particularly by socioeconomic status (SES). While most studies rely on one or two measures of SES to understand these trends, we aim to extend the literature by examining education, occupation, and income in the Swedish context, something not previously studied. Data from the Swedish Longitudinal Occupational Survey of Health (SLOSH) from 2008 to 2020 ( n = 16,497 people age 50 + and n = 64,917 person-observations) were used to examine how each SES measure impacts working life expectancy (WLE). This is a summary measure capturing the average duration that older workers remain in paid work. WLE was estimated using multistate models that assume a continuous-time first-order Markov process, based on transitions between working, not working, and death among older workers. Models were stratified by sex and cross-classified by each SES measure. Findings show that the SES measure used matters for estimating WLE. Specifically, WLE was 1.5 years longer in the highest compared to the lowest education group; 1 year longer in the highest than lowest occupation group, and 0.5 years longer in the highest than lowest income tertiles.