Natural disaster, social cohesion, and prosociality: A natural experiment.

s < .0001), but not a greater likelihood of any volunteering, relative to controls. An instrumental variable analysis supported a causal effect of flood on greater volunteering time and more varied volunteering. Flooding also interacted with social cohesion so that greater volunteering time occurred only among individuals who reported moderate to high levels of cohesion. In addition, prior experience with rescue and recovery after the 9/11 terror attacks robustly predicted prosocial outcomes. Finally, stressful hurricane appraisals fully mediated the effect of flooding on greater volunteering time and more varied helping behavior, suggesting a key positive role for stress appraisals. Plausible confounds, such as residing in the inundation zone, evacuating, and demographic factors, did not account for the flood-prosocial behavior link. Findings indicate that severe disasters can promote more sustained and varied volunteering behavior, that a moderate degree of social cohesion may be necessary for sustained prosociality, and that stress appraisals are potential mechanisms of persistent prosocial responses to severe disaster. (PsycInfo Database Record (c) 2026 APA, all rights reserved).