Rethinking the relative income hypothesis

Income comparisons have been found to be important for individual health. However, the literature has so far looked solely at upward comparisons, disregarding the effects of comparisons with worse-off individuals. In this paper, I use a broad definition of relative income to test simultaneously for the effect of "upward" and "downward" income comparisons on health. Relative deprivation and relative satisfaction indexes are used to summarise upward and downward comparisons. Panel data models are used to correct for income endogeneity bias due to omitted variables. Using German Socio-Economic Panel data (SOEP), results show that relative deprivation has a positive effect, while relative satisfaction has a deleterious impact on health. These findings hold after correcting for unobserved heterogeneity and are robust to using quasi-objective health measures (but mental health) and to different reference groups. -- Relative deprivation ; relative satisfaction ; health