Subjective socioeconomic status and self-rated health in the English Longitudinal Study of Aging: A fixed-effects analysis☆☆We thank the anonymous reviewers of Social Science & Medicine for their helpful comments. Data (Phelps et al., 2020) used in this study as well as programming code for data preparation and analysis (Coustaury et al., 2023) are publicly available. Patrick Präg's work is supported by a grant of the French National Research Agency ANR, ‘Investissements d'Avenir’ (LabEx Ecodec/ANR-11-LABX-0047). The English Longitudinal Study of Aging was developed by a team of researchers based at University College London, Natcen Social Research, the Institute for Fiscal Studies, the University of Manchester, and the University of East Anglia. The data were collected by Natcen Social Research. ELSA funding is currently provided by the National Institute on Aging (Ref: R01AG017644) and by a consortium of UK government departments: Department for Health and Social Care, Department for Transport, Department for Work and Pensions, which is coordinated by the National Institute for Health Research (NIHR, Ref: 198-1074). Funding has also been provided by the Economic and Social Research Council (ESRC)

Higher subjective socio-economic status (SES) goes along with better self-rated health: This finding is well-established in the literature, yet the majority of studies it is based on only rely on cross-sectional analyses and only account for few potential confounders of the association. Particularly wealth, which is increasingly thought of as an important dimension of accumulated advantage, is only rarely examined as a confounder. Using eight waves of panel data from the English Longitudinal Study of Aging (ELSA, 2002–19), we investigate the association between subjective SES and self-rated health. We use random effects models that account for theoretically important time-constant (such as education and social class) and time-varying confounders (such as income and wealth) as well as fixed-effects models, that in addition control for all time-constant confounders, whether observed or unobserved. The fully adjusted fixed-effects model reveals a statistically significant association between subjective SES and self-rated health. Yet, a one-point increase on the subjective SES ladder goes along with a two per cent of a standard deviation increase in self-rated health, only around a quarter of the size of the random-effects estimate. The role of wealth for the subjective SES–self-rated health association is negligible in the fixed-effects specifications. Smoking, drinking, and physical activity do not appear to mediate the association. A substantial part, though not all, of the observed association between subjective SES and self-rated health is due to unobserved confounding rather than a causal effect. Reducing health inequalities based on objective SES is likely more effective than based on subjective SES.