Local economic inequality and depression: Evidence from longitudinal data on local residential contexts and antidepressant use

We study the consequences of economic inequality for depression using individual-level longitudinal administrative data from Denmark (n = 60,654,690 person-time-points). These comprehensive data allow us to i) measure depression (by proxy of redemption of prescriptions for antidepressants) without non-response, ii) measure income inequality at a low level of aggregation to capture individuals’ everyday experiences, iii) conduct within-individual analyses to address confounding from stable individual characteristics, and iv) test whether inequality has similar consequences for people located differently in the local income distribution. In contrast to previous work, we find a modest \textit{negative} average effect of economic inequality (a 1 standard deviation increase in inequality is associated with a 1–2 percent relative decrease in the probability of depression). However, this average masks substantial heterogeneity: the negative effect is confined to the locally relatively well-off, while those with the lowest relative income tend to become more depressed as inequality rises.