BackgroundSocial drift/selection (i.e., mental health symptoms cause low economic status) and social causation theories (i.e., low economic status causes mental health symptoms) specify reciprocal relationships between economic status and mental health. Little is known regarding the disaggregation of within-person and between-person effects in the relationship between economic status and mental health in the long run. The current study sought to examine the reciprocal relationships between economic status and mental health over 20 years of adulthood, disaggregating within-person and between-person effects. MethodsData were from three waves (7108 participants) of the Midlife Development in the United States study. Participants reported information about objective and subjective measures of economic status as well as a wide range of indicators of mental health on the positive dimension, including subjective, social, and psychological well-being, and on the negative side, depression, anxiety, panic attack, anhedonia, somatic amplification, alcohol abuse, and negative affect. Cross-lagged panel models were estimated. ResultsAt between-person levels, both social drift/selection and social causation hypotheses were confirmed when considering subjective measures of economic status. When using objective measures of economic status (i.e., income), the results showed decreased support for social drift/selection and social causation hypotheses. At within-person levels, social drift/selection and social causation hypotheses were generally not supported, with some notable exceptions. LimitationsRestriction of the sample to one country limits the generalizability of the findings. ConclusionsSocial causation and social drift/selection processes act simultaneously mainly at a population level, but much less when considering individual changes. Policy and programs should be targeted at addressing inequality in income and mental health within a nation or a community.