A model of post‐disaster migration responses and income consequences poses that damage severity and individual resilience affect moving decisions. Forced moves are linked to little resilience relative to damage incurred and post‐move income reductions. The empirical analysis analyzes households affected by hurricanes Katrina and Rita. Using American Community Survey data, unobserved heterogeneous income damages are framed as treatment, with the moving decision being the treatment decision. An endogenous switching regression addresses self‐selection issues. The results suggest that movers encountered double victimization: (1) they were forced to move and their income declined; (2) low‐income households were more severely affected than the average.