The Most Vulnerable Suffer Most: How Fear of Unemployment Exacerbates the Mental Health Gradient

Labor market anxiety is a major driver of psychological distress, yet mean effects might obscure significant vulnerabilities among those with low psychological reserves. Therefore, this paper investigates how the impact of job insecurity varies across the mental health distribution. Using German longitudinal data and a quantile regression framework with non-additive fixed effects, we address the endogeneity of fear of job loss by exploiting exogenous firm-specific staff fluctuations as an instrumental variable. Consistent with susceptibility theories, we find that the psychological burden is highly heterogeneous. The estimated penalty is largest at the bottom of the Mental Component Summary (MCS) distribution, reaching 8.8 points at the 10th percentile, and remains substantial between the 20th and 50th percentiles, at approximately 5–6 points. By contrast, the impact is negligible among those with better mental health. We also demonstrate that standard additive quantile fixed effects models mask this gradient, underestimating the burden on vulnerable segments.