Disruption to test scores after hurricanes in the United States.

Abstract

Quantifying how hurricanes disrupt educational attainment is essential to evaluating the burden of climate-related disasters. Here, we examine the association between hurricane-force tropical cyclones and educational attainment among elementary and middle school students in all affected areas in the United States during the 2008/2009-2017/2018 school years. Educational performance was based on county-level average standardized test scores in math and reading/language arts (RLAs). Hurricane-force tropical cyclone-exposed counties were those that experienced a sustained maximal wind speed ⩾64 knots. We estimated the association between hurricane-force tropical cyclone exposure and long-term test scores using a Bayesian hierarchical linear model, accounting for time-varying covariates at the county and grade cohort level. For hurricane-exposed counties, compared with the rest of the state, there were better test scores in Florida in math (β = 0.14; 95% CrI: 0.02, 0.26; PP[β > 0] = 99.0%) and RLA (β = 0.11; 95% CrI: 0.02, 0.22; PP[β > 0] = 99.2%), and worse math scores in North Carolina (β = -0.16; 95% CrI: -0.29, -0.03; PP[β < 0] = 99.4%). Grade cohorts with more racialized and minoritized (e.g. Black, Hispanic, Indigenous) and socioeconomically disadvantaged students tended to have lower test scores, while grade cohorts with greater shares of students racialized as Asian and counties with more college-educated adults tended to have higher scores regardless of hurricane exposure. Disaster preparedness must maximize resilience to climate-related stressors' impacts on academic achievement, especially for vulnerable populations.