Health hazard discrimination or prejudice? A correspondence experiment in Italy.

We study how infectious-disease threats can spill over into discriminatory behavior. Using early COVID-19 in Italy as a case study, we ran an email correspondence experiment with 5356 tourism providers, randomly varying the sender's location and surname to signal origin from areas differentially hit by the first wave. Requests signaling origin from a highly affected area received about 5 percentage points fewer replies and more rejections than observationally equivalent requests; the penalty concentrated on North-sounding surnames and was absent for South-sounding surnames from the same city, pointing to prejudice rather than rational screening on contemporaneous infection risk. While our setting is tourism, the mechanism we uncover-disease-avoidance concerns activating social stereotypes-is general and consistent with theories of social stigma and the behavioral immune system. Such "health-hazard discrimination" can deter testing or travel, undermine equitable access to services, and amplify outbreaks when stigmatized groups avoid contact with providers. We discuss design and policy tools-bias-safe communication, temporary identity-blinding in first contacts, and platform-level fairness nudges-that can mitigate stigma-driven frictions during epidemics. Findings inform preparedness for future outbreaks beyond COVID-19.