Disaster, displacement, and distress: Earthquake intensity and persistent mental health effects in urban New Zealand

Abstract This study estimates the causal effect of spatially varying earthquake intensity on long-term mental health outcomes using linked administrative panel data on 236,856 adults in Greater Christchurch (2007–2019). Exploiting meshblock-level Modified Mercalli Intensity (MMI) variation as a continuous treatment in a difference-in-differences design with individual fixed effects, we find that each unit increase in MMI raises psychiatric medication use by 5.3% (0.051 medications per person-year), with effects persisting six to eight years post-disaster. Impacts are heterogeneous: point estimates are larger for adults aged 65 and older, those without tertiary education, and individuals with severe housing damage, with these differences confirmed by formal tests; effects are also directionally larger for women, though the gender gap itself is not statistically significant. Mediation analysis shows that employment disruption explains the effect among working-age adults but not among the elderly - the most affected group - pointing to unobserved channels such as social network disruption. Hospital admissions rise modestly (5.5%), confirming that exposure primarily generates mild-to-moderate distress. These findings call for differentiated recovery policies targeting labor market support for working-age adults and community-based interventions for older populations.