Causal Effects Between Socioeconomic Status and Health in a Life Course Perspective

Health differences between groups with different socioeconomic status (SES) are a well-established fact but there is debate about the underlying mechanism: Do people have worse health because they are poor or low-educated (social causation) or do they have less education and income because of poor health (health selection)? We study this question using retrospective data from the Survey of Health Aging and Retirement in Europe (SHARE, n≈25,000, 12 countries), the English Longitudinal Study of Ageing (ELSA, n≈7,000), and the US Health and Retirement Study (HRS, n≈8,500). The data include information about events and changes of SES and health from childhood to age at interview (50+). We use structural equation models in a cross-lagged panel design to study the correlated processes of health deterioration and the development of SES. Preliminary findings show, first, that both health and SES depend on their previous status. Second, education is an important mediator between childhood SES and adult SES. Third, health and SES only influence each other at the transition from adulthood to old age. Fourth, social causation is slightly more important than health selection in this transition. Our finding of relatively late mutual effects of SES and health suggest that only later in life when health deteriorates, it seems to be influenced by SES. Likewise only when health shows substantive variation in older age it has a measurable effect on SES. This does not imply an absence of effects in younger ages; it may be due to accumulative effects that can only be identified later in life. We discuss, first, differences between men and women, countries and alternative measures of SES and health, second, conditions for causal interpretation of the findings and, third, implications for policies against health inequalities in aging societies of the finding that both causal mechanisms seem to be at work.