A New Kind of Science

The work assembles a nationwide dataset of hazard losses spanning 1975 to 1998 to assess disaster trends across the United States. It systematically examines temporal patterns of deaths, injuries, and monetary damages by year, and undertakes a spatial assessment of statewide totals. Losses are further normalized by population, land area, and gross domestic product, and an overall hazard score, averaging each state's contribution to national totals of events, deaths, and damages, is used to rank states by proneness.

The analysis moves from describing loss patterns to explaining the disaster-loss burden, surfacing surprises such as North Dakota, Iowa, and Mississippi suffering the greatest per-capita and GDP-relative losses, while states like Florida, Texas, and California rank most prone. Its significance lies in arguing that effective assessment requires developing vulnerability science, creating a national hazard events and losses database, and rethinking how society monitors, assesses, and manages its vulnerabilities for mitigation and planning.