Financial circumstances, mastery, and mental health: Taking unobserved time-stable influences into account.

This paper examines whether low income and subjective financial strain are associated with mental health, as well as whether mastery weakens this association. We analyze three waves of a large sample of Canadians and utilize random and fixed effects regression strategies to assess bias introduced by unobserved time-stable confounders. In random effects models, both low income and subjective financial strain are associated with distress and anger. In fixed effects models that control for all time-stable confounders, the effect of low income is reduced to non-significance for both outcomes. The effect of subjective strain is also reduced in fixed effects models, but remained statistically significant. Sobel tests indicated that the effect of subjective strain on mental health is transmitted through mastery, but this indirect path is modest in magnitude. When interactions are tested, mastery weakens the association between subjective strain and distress, and this effect is robust to the influence of time-stable controls, but mastery does not buffer the subjective strain-anger relationship in either random or fixed-effects models. Finally, moving below the low income threshold increases anger for low mastery individuals, but seems to reduce anger when moving below the low income threshold is coupled with increases in mastery. Collectively, our findings demonstrate the importance of assessing the influence of unobserved time-stable confounders in stress research. Further, discrepancies in the moderating role of mastery reinforce calls for the assessment of multiple outcomes in mental health research.